Stress estimation device, method, and program

Abstract

To accurately estimate the degree of long-term stress of a user that stress is likely to occur at the start time or closing time.SOLUTION: A stress estimation device according to an embodiment comprises a detection unit, an acquisition unit, a calculation unit, a baseline calculation unit, a removal unit, and an estimation unit. The detection unit detects at least one of the start time and closing time for a user who performs work during work hours between the start time and closing time. The acquisition unit acquires biometric waveforms of the user on the basis of a detected result. The calculation unit calculates a short-term stress value from the biometric waveforms. The baseline calculation unit calculates a baseline on the basis of at least the short-term stress value during work hours. The removal unit calculates a stress vibration width of the user by removing the baseline from the calculated short-term stress value. The estimation unit estimates the degree of long-term stress of the user on the basis of the stress vibration width.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] FIELD Embodiments of the present invention relate to a stress estimation device, method, and program. [Background technology]

[0002] Recently, methods for estimating a user's long-term stress level have become known. For example, this type of method calculates an autonomic nervous system index indicating the variability of the user's heart rate (or pulse rate) based on a one-minute heart rate interval (or pulse rate interval) acquired from the user, and estimates the user's long-term stress level based on a short-term stress value determined from the autonomic nervous system index. Because autonomic nervous system indexes vary significantly among individuals, this method assumes that a baseline is established for each individual and that the increase from the baseline corresponds to the level of long-term stress. Specifically, the user is asked to remain calm and rest before starting work. A first short-term stress value before starting work is used as the baseline, and the increase in stress from the baseline is measured as a response to the workload. That is, this method calculates a first short-term stress value before starting work and a second short-term stress value during work. Then, by subtracting the first short-term stress value (baseline) from the second short-term stress value after completing work, the level of long-term stress is estimated based on the increase from the baseline.

[0003] However, according to the inventor's research, the above-described method is difficult to apply to office workers, among various types of users. For example, office workers often experience delays in work progress or transportation delays, making it difficult to remain calm and restful at the start of work. In this case, their heart rate (or pulse rate) increases before work begins, raising the first short-term stress value (baseline), and decreasing the estimated long-term stress. Meanwhile, other office workers may experience delays in work progress or the inability to work overtime, making it difficult to remain calm and restful at the end of work after completing work. In this case, their heart rate (or pulse rate) increases after work is completed, raising the second short-term stress value and increasing the estimated long-term stress. In other words, for users such as office workers, stress tends to occur at the start or end of work, causing the estimated long-term stress to fluctuate, making the above-described method difficult to apply. Therefore, it is desirable to be able to appropriately estimate the magnitude of long-term stress for users who tend to experience stress at the start or end of work. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 7048709 Summary of the Invention [Problem to be solved by the invention]

[0005] The problem to be solved by the present invention is to provide a stress estimation device, method, and program that can appropriately estimate the magnitude of long-term stress for a user who is prone to experiencing stress at the start or end of work. [Means for solving the problem]

[0006] A stress estimation device according to an embodiment includes a detection unit, an acquisition unit, a calculation unit, a baseline calculation unit, a removal unit, and an estimation unit. The detection unit detects at least one of the start and end of work for a user who works between the start and end of work. The acquisition unit acquires the user's biowaveforms based on the detection results. The calculation unit calculates a short-term stress value from the biowaveforms. The baseline calculation unit calculates a baseline based on at least the short-term stress value during work. The removal unit calculates the user's stress fluctuation by removing the baseline from the calculated short-term stress value. The estimation unit estimates the magnitude of the user's long-term stress based on the stress fluctuation. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a stress estimation device according to a first embodiment. [Figure 2] 4 is a flowchart for explaining the operation in the first embodiment. [Figure 3] FIG. 3 is a schematic diagram for explaining the operation of the first embodiment. [Figure 4] FIG. 10 is a block diagram showing an example of the configuration of a stress estimation device according to a second embodiment. [Figure 5] 10 is a flowchart for explaining the operation in the second embodiment. [Figure 6] FIG. 10 is a block diagram showing an example of the configuration of a stress estimation device according to a third embodiment. [Figure 7] FIG. 11 is a schematic diagram for explaining a regression model dictionary according to the third embodiment. [Figure 8] 10 is a flowchart for explaining the operation in the third embodiment. [Figure 9] FIG. 10 is a schematic diagram for explaining a modified example of the third embodiment. [Figure 10] FIG. 10 is a schematic diagram for explaining a modified example of the third embodiment. [Figure 11] FIG. 11 is a block diagram showing an example of the configuration of a stress estimation device according to another modified example of the third embodiment. [Figure 12] 10 is a flowchart for explaining the operation of another modified example of the third embodiment. [Figure 13] FIG. 10 is a schematic diagram for explaining the operation of the fourth embodiment. [Figure 14] FIG. 10 is a schematic diagram for explaining a modified example of the fourth embodiment. [Figure 15] FIG. 13 is a block diagram showing an example of a hardware configuration of a stress estimation device according to a fifth embodiment. [Figure 16] FIG. 13 is a block diagram showing an example of a hardware configuration of a stress estimation system including a stress estimation device according to a modified example of the fifth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, a stress estimation device, a method, and a program according to each embodiment will be described with reference to the drawings. In each embodiment, parts with the same reference numerals perform similar operations, and redundant explanations will be omitted as appropriate.

[0009] First Embodiment 1 is a block diagram showing an example of the configuration of a stress estimation device according to embodiment 1. The stress estimation device 10 includes an employment change detection unit 1, a waveform acquisition unit 2, a short-term stress calculation unit 3, a baseline accumulation unit 5, a baseline removal unit 6, and a stress estimation unit 7.

[0010] Here, the work change detection unit 1 detects at least one of the start time and the end time of a workday for a user who works between the start time and the end time of the workday. For example, the work change detection unit 1 may detect the start time of a workday based on a user's startup operation, and may detect the end time of a workday based on a user's shutdown operation. The target of the startup operation and the shutdown operation may be a user's personal computer (PC) or a login app on the user's PC. That is, the work change detection unit 1 may be installed in a user's PC, which is an office device used by the user for work. In this case, the work change detection unit 1 may detect at least one of the start time and the end time of a workday based on the user's operation on the office device used for work. Here, the office device refers to a device dedicated to the user in the office, such as a user's PC. Additionally, the office device does not refer to a shared device in the office, such as a printer, copier, or shredder. Furthermore, when detecting the start time of a workday, the work change detection unit 1 may distinguish between a morning start time and an afternoon start time based on the startup operation and the time of day. For example, the start time of an afternoon shift is detected when there are two start times in a day, such as the start of an afternoon shift (after lunch break) after a morning holiday, or a morning and evening shift. Furthermore, the working hours are not detected because they are obtained as a certain period of time after the start time is detected. The work change detection unit 1 is an example of a detection unit.

[0011] The waveform acquisition unit 2 acquires the user's biometric waveform based on the detection results. For example, if the work change detection unit 1 detects the start of work, the waveform acquisition unit 2 acquires the biometric waveform at the detected start of work. Alternatively, if the work change detection unit 1 detects the end of work, the waveform acquisition unit 2 acquires the biometric waveform at the detected end of work. Specifically, the waveform acquisition unit 2 includes a sensor 21, an ADC 22, and a waveform storage unit 23. ADC stands for analog-digital converter. The sensor 21 detects biometric data from the user. For example, at least one of a pulse waveform or an electrocardiogram waveform can be used as the biometric data. The ADC 22 converts the analog data output from the sensor 21 into digital data and outputs the digital data. For example, the pulse waveform or the electrocardiogram waveform is converted into digital data at a sampling interval of 1 ms (sampling frequency 1 kHz). The waveform storage unit 23 stores the digital data in chronological order. The stored digital data corresponds to a biometric waveform representing the biometric data in chronological order. Additionally, the waveform acquisition unit 2 may be configured to acquire a biometric waveform from a user using a biometric sensor. The biometric sensor may include, for example, at least one of an electrocardiogram sensor, a photoplethysmogram sensor, a chest band sensor, a wristband sensor, a mouse-mounted sensor, and an optical camera. A mouse-mounted sensor may be, for example, a photoplethysmogram sensor mounted on the mouse of a user's PC at the position where the user's fingertip touches the mouse. In this case, only biometric waveforms that are stably collected by maintaining stable contact with the user's fingertip may be stored. The optical camera may be mounted on, for example, a smartphone, and may capture video data of the user's fingertip or face, and acquire a biometric waveform of a pulse waveform corresponding to changes in the color of the fingertip or facial color from changes in the green component of the video data. The chest band, wristband, and smartphone in the biometric sensor are examples of devices that can be carried by the user. That is, the waveform acquisition unit 2 may be disposed in a device that can be carried by the user. However, the present invention is not limited to this, and the waveform acquisition unit 2 may be arranged in an office device used by a user at work, such as when an optical camera is mounted on a user's PC.The waveform acquisition unit 2 is an example of an acquisition unit.

[0012] The short-term stress calculation unit 3 calculates a short-term stress value from the biometric waveform. The short-term stress calculation unit 3 calculates the RR interval (RRI, heartbeat interval), which is the interval between heartbeats, from the electrocardiogram waveform, or the pulse wave peak-to-peak interval (PPI, pulse interval), which is the interval between peaks, from the pulse waveform, and calculates various short-term stress values ​​by analyzing fluctuations in the RRI or PPI. For example, the short-term stress calculation unit 3 calculates a short-term stress value at the start of work from the biometric waveform at the start of work. Alternatively, the short-term stress calculation unit 3 calculates a short-term stress value at the end of work from the biometric waveform at the end of work. Furthermore, autonomic nervous indices related to fluctuations in heartbeat intervals and pulse intervals can be used as short-term stress values. For example, RR_mean, RR_std, CV, NN50, pNN50, NN40, pNN40, NN30, pNN30, RMSSD, LF, HF, and LF / HF can be used as appropriate. Here, RR_mean is the mean heartbeat interval, representing the average over a certain section of the heartbeat interval. RR_std is the standard deviation of the heartbeat interval and represents the standard deviation over a certain interval of the heartbeat interval. CV is the coefficient of variation of the heartbeat interval and represents the coefficient of variation (standard deviation / average) over a certain interval of the heartbeat interval. NN50 represents the number of intervals where the difference between adjacent heartbeat intervals is 50 ms or more. pNN50 represents the percentage of NN50 intervals. NN40 represents the number of intervals where the difference between adjacent heartbeat intervals is 40 ms or more. pNN40 represents the percentage of NN40 intervals. NN30 represents the number of intervals where the difference between adjacent heartbeat intervals is 30 ms or more. pNN30 represents the percentage of NN30 intervals. RMSSD is an abbreviation for Root Mean Square of Successive Differences and represents the square root of the average of the squares of the differences between consecutively adjacent heartbeat intervals. LF is an abbreviation for Low Frequency and represents the low-frequency component of the heartbeat interval (0.04 to 0.15 Hz). HF is an abbreviation for High Frequency and represents the high frequency component (0.15 to 0.40 Hz) of the heartbeat interval. LF / HF is the ratio of LF to HF and represents the balance between the degree of sympathetic nervous tension (LF) and the degree of parasympathetic nervous tension (HF). In other words, various autonomic nervous indexes in the frequency domain or time domain can be used as short-term stress values.Here, the fixed interval refers to the time span to be analyzed, for example, 60 [s]. In this case, the short-term stress value is calculated at intervals of at least 60 [s]. Here, the calculation method for various short-term stress values ​​has been explained using heartbeat intervals as input, but the calculations are similar even when pulse intervals are used as input. Various known methods can be applied to calculate the short-term stress value. The short-term stress calculation unit 3 is an example of a calculation unit.

[0013] The baseline accumulation unit 5 calculates a baseline based on at least the short-term stress value during work hours and accumulates the baseline in memory. Here, work hours, for example, begin a certain time after the start of work. Additionally, work hours are switched from the start of work hours depending on a predetermined time period after the start of work hours, regardless of the work content during work hours. The baseline may also be the minimum value among the short-term stress value at the start of work hours, the short-term stress value during work hours, and the short-term stress value at the end of work hours. The baseline may also be the minimum value among the average short-term stress value at the start of work hours, the average short-term stress value during work hours, and the average short-term stress value at the end of work hours. In any case, the short-term stress value during work hours is taken into consideration when calculating the baseline. The baseline accumulation unit 5 is an example of a baseline calculation unit.

[0014] The baseline removal unit 6 calculates the user's stress fluctuation by removing the baseline from the calculated short-term stress value. For example, the baseline removal unit 6 calculates the stress fluctuation at the start of work by removing the baseline from the short-term stress value at the start of work. Specifically, the stress estimation device 10 typically performs stress measurement at the start of work, when short-term stress is likely to appear, and uses the long-term average or minimum value of the short-term stress value measured during work as the baseline for each individual's short-term stress value to eliminate individual differences. The long-term average or minimum value of the short-term stress value during work reflects the period of calm and rest during work. The baseline removal unit 6 calculates the difference (average fluctuation at the start of work) obtained by removing the baseline from the short-term stress value at the start of work as the stress fluctuation at the start of work. For example, the baseline removal unit 6 calculates the stress fluctuation at the end of work by removing the baseline from the short-term stress value at the end of work. As described above, the stress fluctuation at the end of work refers to the average fluctuation at the end of work. The stress fluctuation range at the start of work and the stress fluctuation range at the end of work are each used to estimate the stress value on a daily basis. The baseline removal unit 6 is an example of a removal unit.

[0015] The stress estimation unit 7 estimates the user's long-term stress level based on the stress fluctuation. Daily stress fluctuations are input, and the estimated long-term stress level is output daily. The long-term stress level can be measured using subjective assessment values, such as a subjective questionnaire, which score various user responses to multiple questions. Examples of subjective assessment values ​​that can be used include the 21-item BDI-II, which assesses depression over the past two to three days; the 20-item Self-Defensive Scale (SDS), which assesses current depression; the 40-item STAI (State-Trait Anxiety Inventory), which assesses current and usual feelings; the 57-item Occupational Stress Questionnaire, which assesses the user's state of mind over the past month; the 65-item POMS-2, which assesses emotions over the past week; and the 10-item PSS-10, which assesses emotions and behavior over the past month. BDI-II is an abbreviation for the Beck Depression Inventory-Second Edition. SDS is an abbreviation for Self-rating Depression Scale. STAI is an abbreviation for State-Trait Anxiety Inventory. POMS2 is an abbreviation for Profile of Mood States 2nd Edition. PSS is an abbreviation for Perceived Stress Scale. For example, the stress estimation unit 7 may have a function that calculates a subjective assessment value from stress fluctuations based on objective biometric waveforms, and the magnitude of long-term stress, which is a subjective assessment value, may be estimated from the stress fluctuations using this function. As the function, for example, a regression model (regression equation) that expresses the relationship between stress fluctuations, which is an explanatory variable, and the magnitude of long-term stress, which is a dependent variable, may be used as appropriate. This regression model is created in advance based on the relationship between stress fluctuations and the magnitude of long-term stress in an unspecified number of individuals. Additionally, the regression model is a model that expresses the relationship between an objective value (stress fluctuations), such as an autonomic nervous system index, and a subjective value (the magnitude of long-term stress), such as a score on a subjective questionnaire. The regression model may be a simple regression model or a multiple regression model, a linear regression model, or a nonlinear regression model.Alternatively, the stress estimation unit 7 may be a model using machine learning such as Lasso regression, Ridge regression, ElasticNet regression, LightGBM, or Random Forest regression. The regression model may also be called a regression equation. The stress estimation unit 7 is an example of an estimation unit.

[0016] Next, the operation of the stress estimation device configured as above will be described with reference to the flowchart of FIG. 2 and the schematic diagram of FIG.

[0017] (Step ST10) The baseline accumulation unit 5 calculates and accumulates a baseline of the short-term stress value from the user's biowaveform in advance. Specifically, the baseline accumulation unit 5 calculates the baseline based on at least the short-term stress value during work among the short-term stress values ​​calculated from the biowaveform at the start of work, during work, and at the end of work, and accumulates the baseline in memory. In this example, the short-term stress value is assumed to be RR_mean (average heartbeat interval). Furthermore, the baseline is assumed to be the minimum value of RR_mean at the start of work, RR_mean during work, and RR_mean at the end of work.

[0018] (Step ST20) The work change detection unit 1 detects at least one of the start time and the end time of a workday for a user who works during a shift between the start time and the end time of a workday. For example, the work change detection unit 1 detects the start time of a workday by the user's login application startup operation.

[0019] (Step ST30) The waveform acquiring unit 2 acquires the user's biowaveform based on the result of detection in step ST20. For example, the waveform acquiring unit 2 acquires the detected electrocardiogram waveform at the start of work as the biowaveform.

[0020] (Step ST40) The short-term stress calculation unit 3 calculates a short-term stress value from the biological waveform acquired in step ST30. For example, the short-term stress calculation unit 3 calculates a short-term stress value at the start of work from the biological waveform at the start of work. Specifically, for example, the short-term stress calculation unit 3 calculates RR_mean (average heartbeat interval) from the electrocardiogram waveform at the start of work as the short-term stress value.

[0021] (Step ST60) The baseline removal unit 6 calculates the user's stress fluctuation by removing the baseline from the short-term stress value calculated in step ST40. In this example, the stress fluctuation represents the fluctuation range of RR_mean (heart rate variability) at the start of work.

[0022] (Step ST70) The stress estimation unit 7 estimates the user's long-term stress level based on the stress fluctuation calculated in step ST60. For example, as shown in FIG. 3, the stress estimation unit 7 performs the estimation using a regression model y=ax+b, which represents the relationship between the stress fluctuation level x and the long-term stress level y. For ease of explanation, a simple linear regression is used, where a represents the regression coefficient and b represents the y-intercept. The stress estimation unit 7 may also estimate a user whose estimated long-term stress level exceeds a threshold as a mentally ill person (high-stress person), or may estimate a user whose estimated long-term stress level is below the threshold as a healthy person (low-stress person). In either case, the stress estimation unit 7 outputs the estimation result, including the long-term stress level, to a display or the like. Thereafter, the stress estimation device 10 terminates the processing.

[0023] As described above, according to the first embodiment, the work change detection unit 1 detects at least one of the start and end times of a workday for a user who works between the start and end times. The waveform acquisition unit 2 acquires the user's biowaveform based on the detection results. The short-term stress calculation unit 3 calculates a short-term stress value from the biowaveform. The baseline accumulation unit 5 calculates a baseline based on at least the short-term stress value during work. The baseline removal unit 6 calculates the user's stress fluctuation by removing the baseline from the calculated short-term stress value. The stress estimation unit 7 estimates the user's long-term stress level based on the stress fluctuation. In this way, by removing the baseline based on at least the short-term stress value during work from the short-term stress value during work, the long-term stress level can be appropriately estimated for users who tend to experience stress during work. Furthermore, unlike conventional systems, the first embodiment does not use short-term stress values ​​during workdays, which tend to manifest as stress, as the baseline, and instead suppresses increases or decreases in the estimated long-term stress, thereby enabling appropriate estimation of the long-term stress level. Furthermore, by removing the user's baseline from the user's short-term stress value, it is possible to expect an improvement in the accuracy of the stress fluctuation range.

[0024] Furthermore, according to the first embodiment, the work change detection unit 1 detects the start of work. The waveform acquisition unit 2 acquires the detected bio-waveform at the start of work. The short-term stress calculation unit 3 calculates the short-term stress value at the start of work from the bio-waveform at the start of work. The baseline removal unit 6 calculates the stress fluctuation range at the start of work by removing the baseline from the short-term stress value at the start of work. Therefore, similar to the effect described above, the magnitude of the user's long-term stress at the start of work can be appropriately estimated.

[0025] Furthermore, according to the first embodiment, the work change detection unit 1 may detect the end of work. When the work change detection unit 1 detects the end of work, the waveform acquisition unit 2 acquires the biowaveform at the detected end of work. The short-term stress calculation unit 3 calculates the short-term stress value at the end of work from the biowaveform at the end of work. The baseline removal unit 6 calculates the stress fluctuation range at the end of work by removing the baseline from the short-term stress value at the end of work. Therefore, in this case, similar to the effect described above, the magnitude of the user's long-term stress at the end of work can be appropriately estimated.

[0026] Furthermore, according to the first embodiment, the baseline is the minimum of the short-term stress value at the start of work, the short-term stress value during work, and the short-term stress value at the end of work. Therefore, in addition to the effects described above, the baseline can be the short-term stress value when the user is most calm and at rest during the entire period from the start of work, work, and end of work. Therefore, even if the user has difficulty in remaining calm and at rest, an appropriate baseline can be calculated.

[0027] Furthermore, according to the first embodiment, the baseline may be the minimum of the average short-term stress values ​​at the start of work, the average short-term stress values ​​during work hours, and the average short-term stress values ​​at the end of work hours. In this case, in addition to the effects described above, when one short-term stress value becomes the minimum value (abnormal value) due to some abnormality, the minimum value of the average values ​​calculated including the abnormal value is used as the baseline, thereby mitigating the influence of the abnormal value.

[0028] Furthermore, according to the first embodiment, the work change detection unit 1 detects the start time of work through a start operation by the user. Alternatively, the work change detection unit 1 may detect the end time of work through a stop operation by the user. Therefore, in addition to the effects described above, the start time and end time of work can be detected appropriately. To add to this, when the start time of work is detected simply based on the time, there is the inconvenience that the start time of work cannot be detected when the user arrives at work during a time period that does not include the time used for detection, such as when the user is on a morning holiday or is late. In contrast, according to the first embodiment, the start time of work is detected through a start operation by the user, so the start time of work can be detected appropriately. The same applies to the detection of the end time of work.

[0029] Furthermore, according to the first embodiment, when detecting the start time of work, the work change detection unit 1 may distinguish between the morning start time and the afternoon start time based on the activation operation and the time. In this case, the above-described effects can be obtained even for work patterns that have an afternoon start time, such as workplaces where employees start work in the afternoon after a morning holiday, or workplaces where employees start work twice a day, such as morning and evening shifts.

[0030] <Second embodiment> Next, a stress estimation device according to a second embodiment will be described.

[0031] The second embodiment is a modification of the first embodiment, and represents a specific example of switching the processing of short-term stress values ​​between the start of work and the end of work.

[0032] Fig. 4 is a block diagram showing an example of the configuration of a stress estimation device according to the second embodiment. Compared to the configuration shown in Fig. 1, this stress estimation device 10 has a work start detection unit 1a and a baseline calculation unit 5a instead of the work change detection unit 1 and the baseline accumulation unit 5, and further includes a work timing determination unit 4.

[0033] Here, the work start detection unit 1a is a specific example of the work change detection unit 1, and detects the start time of a user who works between the start time and the end time of the work. The method of detecting the start time by the work start detection unit 1a is the same as the method by the work change detection unit 1.

[0034] The work timing determination unit 4 sends the short-term stress value calculated by the short-term stress calculation unit to at least one of the baseline calculation unit 5a and the baseline removal unit 6 based on the results detected by the work start detection unit 1a. For example, the work timing determination unit 4 sends the short-term stress value at the start of work to the baseline removal unit 6 based on the results of detecting the start of work. Furthermore, the work timing determination unit 4 determines the start timing of work depending on whether a certain amount of time has passed since the start of work was detected, and sends the short-term stress value at work to the baseline calculation unit 5a. The work timing determination unit 4 is an example of a sending unit.

[0035] The baseline calculation unit 5a is a specific example of the baseline storage unit 5, and calculates a baseline based on the short-term stress value during work that is sent out, and stores the baseline in memory.

[0036] The baseline removal unit 6 calculates the user's stress fluctuation range by removing the baseline from the transmitted short-term stress value.

[0037] The other configurations are the same as those in the first embodiment.

[0038] Next, the operation of the stress estimation device configured as above will be described with reference to the flowchart in Fig. 5. In Fig. 5, step ST50, enclosed by a dashed line, is added between step ST40 and step ST60 shown in Fig. 2. In addition, in this embodiment, step ST10, shown in Fig. 2, for storing a baseline in advance is omitted.

[0039] First, step ST20 is executed in the same manner as described above, and the work start detection unit 1a detects the work start time of a user who works during a shift between the start time and the end time of the work day.

[0040] Similarly, steps ST30 to ST40 are executed to acquire the user's biological waveform and calculate a short-term stress value from the biological waveform. Thereafter, step ST50 is executed. Step ST50 includes steps ST51 to ST55.

[0041] (Step ST50) The work timing determination unit 4 determines whether the work start detection unit 1a has detected the start of work (step ST51). If the result of this determination is that the work start has been detected, the work timing determination unit 4 sends the short-term stress value at the start of work to the baseline removal unit 6 (step ST52), and proceeds to step ST60.

[0042] On the other hand, if the result of the determination in step ST51 is No, the work timing determination unit 4 determines the start timing of work depending on whether a certain time has passed since the timing at which the start of work was detected (step ST53).If the result of this determination is No, the process returns to step ST51.

[0043] If the result of the determination in step ST53 is that it is the start timing of work, the work timing determination unit 4 sends the short-term stress value at work to the baseline calculation unit 5a (step ST54).

[0044] After step ST54, the baseline calculation unit 5a calculates a baseline based on the short-term stress value during work that has been sent, and stores the baseline in memory (step ST55). Upon completion of step ST52 or ST55, step ST50 ends.

[0045] Thereafter, the processing from step ST60 onwards is executed in the same manner as described above. In the second embodiment, the baseline removal unit 6 calculates the user's stress fluctuation range at the start of work by removing the baseline during work from the short-term stress value at the start of work (step ST60).

[0046] The stress estimation unit 7 estimates the magnitude of the user's long-term stress based on the stress fluctuation range at the start of work (step ST70), and outputs the estimation result to a display or the like.

[0047] As described above, according to the second embodiment, the work timing determination unit 4 sends the short-term stress value to at least one of the baseline calculation unit 5a and the baseline removal unit 6 based on the results detected by the work start detection unit 1a. Therefore, in addition to the effects described above, the configuration in which the most recent short-term stress value is used for baseline calculation and baseline removal makes it possible to estimate the level of the user's long-term stress based on the most recent baseline.

[0048] <Third embodiment> Next, a stress estimation device according to a third embodiment will be described.

[0049] The third embodiment is a modification of the second embodiment, and represents a specific example of using a regression model or threshold determination when estimating stress.

[0050] 6 is a block diagram showing an example of the configuration of a stress estimation device 10 according to the third embodiment. Compared to the configuration shown in FIG. 4, this stress estimation device 10 further includes a regression model dictionary 8 and a threshold determination unit 9.

[0051] Here, the regression model dictionary 8 stores a regression model representing the relationship between stress fluctuation and long-term stress, as shown in Figure 3. Specifically, the regression model is created from the results of a preliminary examination of the relationship between stress fluctuation and long-term stress for an unspecified number of subjects. The long-term stress level can be obtained as a subjective assessment value (e.g., PSS-10 score) for an unspecified number of subjects, as shown in Figure 7. In Figure 7, the vertical axis represents the PSS-10 score, and the horizontal axis represents the subject number. PSS-10 scores range from 0 to 40, with a score below 14 representing low stress, a score between 14 and 27 representing moderate stress, and a score above 27 representing high perceived stress. Note that the subjective assessment value is not limited to the PSS-10 score; as mentioned above, scores from the BDI-II, SDS, STAI, the Brief Occupational Stress Questionnaire, POMS2, and other instruments can also be used as appropriate. The stress fluctuation is an objective assessment value of the same type as the short-term stress value described above. As described above, RR_mean, RR_std, CV, NN50, pNN50, NN40, pNN40, NN30, pNN30, RMSSD, LF, HF, and LF / HF can be used as appropriate for stress fluctuation and short-term stress type. For example, the regression model dictionary 8 stores a regression model in advance for each type of stress fluctuation. That is, the regression model dictionary 8 stores regression models associated with the type of stress fluctuation, such as storing a first regression model associated with the pulse wave peak interval and a second regression model associated with RR_mean. The regression model dictionary 8 is an example of a storage unit.

[0052] Accordingly, the stress estimation unit 7 estimates the magnitude of the user's long-term stress based on a regression model in the regression model dictionary 8. The stress estimation unit 7 also estimates the magnitude of the user's long-term stress based on a regression model corresponding to the type of calculated stress fluctuation range.

[0053] The threshold determination unit 9 outputs an alert when the magnitude of long-term stress estimated by the stress estimation unit 7 exceeds a threshold. For example, in FIG. 3, when the magnitude of long-term stress is a PSS-10 score, 27 points can be used as the threshold. However, the threshold is not limited to this, and other scores that are appropriate as thresholds even in the case of a PSS-10 score (for example, 23 points for detecting the risk of early-stage long-term stress) may be used. Also, for example, the threshold determination unit 9 compares the estimated magnitude of long-term stress with a threshold and outputs an alert depending on the comparison result. The threshold determination unit 9 is an example of an output unit.

[0054] Next, the operation of the stress estimation device configured as above will be described with reference to the flowchart of Fig. 8. In Fig. 8, step ST70 enclosed by a dashed line is executed as a specific example of step ST70 of the second embodiment.

[0055] Now, it is assumed that steps ST20 to ST60 are executed in the same manner as described above, and the user's stress fluctuation range at the start of work is calculated in step ST60.

[0056] (Step ST70) In step ST70, the stress estimation unit 7 estimates the magnitude of the user's long-term stress based on the stress fluctuation range at the start of work. Step ST70 includes steps ST71 to ST77.

[0057] The stress estimation unit 7 acquires a regression model from the regression model dictionary 8 based on the type of stress fluctuation (step ST71). The stress estimation unit 7 also estimates the magnitude of long-term stress based on the calculated stress fluctuation and the acquired regression model (step ST72), outputs the estimation result to a display or the like, and stores the magnitude of long-term stress in memory in association with a date.

[0058] After step ST72, the stress estimation unit 7 determines whether the magnitude of the long-term stress for several days falls within a predetermined range (step ST73), and if not, estimates the estimation result of step ST72 as transient stress (step ST74). If the determination result of step ST73 indicates that the magnitude falls within the predetermined range, estimates the estimation result of step ST72 as chronic stress (step ST75). The stress estimation unit 7 also outputs the estimation result of step ST74 or ST75 to a display or the like, and proceeds to step ST76.

[0059] In step ST76, the threshold determination unit 9 determines whether the magnitude of the long-term stress estimated in step ST72 exceeds the threshold, and if not, ends the process. If the result of the determination in step ST76 is that the threshold is exceeded, the threshold determination unit 9 outputs an alert to a display or the like (step ST77), and ends the process. Note that steps ST76 to ST77 may be executed between steps ST72 and ST73.

[0060] As described above, according to the third embodiment, the regression model dictionary 8 stores in advance a regression model that represents the relationship between stress fluctuation and the magnitude of long-term stress. Therefore, in addition to the effects described above, the magnitude of long-term stress can be easily estimated by using the pre-stored regression model.

[0061] Furthermore, according to the third embodiment, the regression model dictionary 8 stores a regression model in advance for each type of stress fluctuation. The stress estimation unit 7 estimates the magnitude of the user's long-term stress based on the regression model corresponding to the calculated type of stress fluctuation. Therefore, in addition to the effects described above, the magnitude of long-term stress can be estimated from various types of stress fluctuation using the regression model corresponding to the type, thereby improving versatility.

[0062] Furthermore, according to the third embodiment, the threshold determination unit 9 outputs an alert when the magnitude of the estimated long-term stress exceeds a threshold. Therefore, in addition to the effects described above, the alert can encourage users with high long-term stress to take measures such as resting.

[0063] Furthermore, according to the third embodiment, if the magnitude of long-term stress over several days falls within a predetermined range, it is possible to detect chronic stress rather than transient stress. Additionally, since the stress estimation device 10 can accurately estimate the stress value on a daily basis, it can observe the daily changes and detect, for example, chronic stress that continues for two weeks or more (chronic stress), thereby detecting mental stress-related illnesses at an early stage and encouraging industrial physicians to intervene early.

[0064] <Modification of the third embodiment> (First Modification) Although the third embodiment uses a regression model that represents the relationship between stress fluctuations based on one type of short-term stress value and the magnitude of long-term stress, this is not limiting. For example, a first modification of the third embodiment uses a regression model that represents the relationship between stress fluctuations based on three types of short-term stress values ​​and the magnitude of long-term stress. Here, the stress fluctuations based on the three types of short-term stress values ​​may be generated, for example, by acquiring three types of biowaveforms from each of an unspecified number of subjects, calculating three types of short-term stress values, removing the baseline from each of the three types of short-term stress values, and combining the three types of stress fluctuations. As shown in an example in Figure 9, the stress fluctuations based on the three types of short-term stress values ​​correspond to the magnitude of long-term stress (PSS-10 score). In Figure 9, the vertical axis represents the PSS-10 score, and the horizontal axis represents the stress fluctuations based on the three types of short-term stress values. There is no multicollinearity between the three types of short-term stress values. Each point in Figure 9 corresponds to one of 80 unspecified number of subjects. In addition, in Figure 9, the relationship between the stress fluctuation on the horizontal axis and the PSS-10 score on the vertical axis has a (Spearman's rank) correlation coefficient of 0.453 with a multiple regression curve using three types of stress fluctuation, indicating a positive correlation. Also in Figure 9, a PSS-10 score of 23 or more is considered to be a high-stress individual, and a score below 23 is considered to be a healthy individual. Here, the results of binary classification when the threshold score is 23 are expressed as an ROC curve showing the relationship between the false positive rate (FPR) and the true positive rate (TPR), as shown in an example in Figure 10. In Figure 10, the horizontal axis is the false positive rate (FPR), and the vertical axis is the true positive rate (TPR). The false positive rate (FPR) and true positive rate (TPR) can be obtained as shown in the following equations.

[0065] TPR = number of correctly predicted high stress individuals / number of high stress individuals FPR = number of low-stress people incorrectly predicted as high-stress people / number of low-stress people ROC is an abbreviation for Receiver Operating Characteristic. A higher true positive rate (TPR) and a lower false positive rate (FPR) are better for the ROC curve, so the larger the area under the ROC curve (ROC-AUC), the better. In Figure 10, the ROC-AUC is an estimation accuracy of 0.806. The diagonal dashed line in Figure 10 represents the case where the ROC-AUC is 0.5.

[0066] According to the first modification of the third embodiment described above, in addition to the above-mentioned effects, by increasing the number of types of stress fluctuations used to estimate the magnitude of long-term stress, it is possible to expect improved estimation accuracy. Furthermore, although the first modification uses stress fluctuations based on three types of short-term stress values ​​in the regression model, stress fluctuations based on any multiple types of short-term stress values ​​can also be used in the regression model, not limited to three types. However, there is no multicollinearity between the multiple types of short-term stress values.

[0067] (Second Modification) In the third embodiment, the work start detection unit 1a detects the start of work, and the work timing determination unit 4 transmits the short-term stress value at the start of work to the baseline removal unit 6. However, this is not limiting. For example, as shown in FIG. 11 , the work start detection unit 1a may be replaced with the work change detection unit 1 described above. Accordingly, the work timing determination unit 4 transmits the short-term stress value calculated by the short-term stress calculation unit based on the results detected by the work change detection unit 1 to at least one of the baseline calculation unit 5a and the baseline removal unit 6. For example, the work timing determination unit 4 transmits the short-term stress value at the start of work to the baseline removal unit 6 based on the results detected by the work start. Furthermore, the work timing determination unit 4 determines the start timing of work depending on whether a certain amount of time has passed since the detection of the start of work, thereby transmitting the short-term stress value at work to the baseline calculation unit 5a. Furthermore, the work timing determination unit 4 transmits the short-term stress value at the end of work to the baseline removal unit 6 based on the results detected by the work end.

[0068] The baseline removal unit 6 calculates the user's stress fluctuation by removing the baseline from the transmitted short-term stress value. For example, the baseline removal unit 6 calculates the user's stress fluctuation at the end of work by removing the baseline from the transmitted short-term stress value at the end of work.

[0069] The other configurations are the same as those of the third embodiment.

[0070] Next, the operation of the stress estimation device configured as described above will be described using the flowchart in Fig. 12. In the following description, the "start-of-work detection unit 1a" in the description using Fig. 5 will be replaced with the "work change detection unit 1" as appropriate, using the phrase "similar to the above," to avoid repetitive explanation. In Fig. 12, steps ST56 and ST57 on the right side are added during the return from step ST53 to step ST51 in step ST50 shown in Fig. 5.

[0071] That is, it is assumed that steps ST20 to ST51 have been executed in the same manner as described above. If the result of the determination in step ST51 is that the start of work has been detected, the process proceeds to step ST60 via step ST52, as described above.

[0072] If the result of the determination in step ST51 is negative, the determination in step ST53 is made in the same manner as described above. If the result of the determination is that it is the start timing of work, the process proceeds to step ST60 via steps ST54 and ST55.

[0073] On the other hand, if the result of the determination in step ST53 is No, the work timing determination unit 4 determines whether the end of work has been detected by the work change detection unit 1 (step ST56). If the result of this determination is No, the process returns to step ST51. If the result of the determination in step ST56 is that the end of work has been detected, the work timing determination unit 4 sends the short-term stress value at the time of the end of work to the baseline removal unit 6 (step ST57), and proceeds to step ST60.

[0074] Thereafter, the processes from step ST60 onward are executed in the same manner as described above. In the second modified example, the baseline removal unit 6 calculates the user's stress fluctuation range at the end of work, for example, by removing the baseline during working hours from the short-term stress value at the end of work (step ST60).

[0075] The stress estimation unit 7 estimates the magnitude of the user's long-term stress based on the stress fluctuation range at the end of work (step ST70), and outputs the estimation result to a display or the like.

[0076] According to the second modification of the third embodiment described above, in addition to the effects of the third embodiment, it is possible to appropriately estimate the magnitude of long-term stress for a user who is prone to experiencing stress at the end of work hours.

[0077] (Third Modification) In the third embodiment, it is determined whether the magnitude of long-term stress for several days falls within a predetermined range, and if so, chronic stress is estimated. However, this is not limiting. For example, chronic stress may be estimated based on the magnitude of long-term stress for several days by utilizing a machine learning model using a neural network such as CNN or LSTM. This third modification can also achieve the same effects as the third embodiment.

[0078] (Fourth Modification) In the fourth embodiment, it is determined whether the magnitude of the estimated long-term stress exceeds a threshold, and an alert is output when the threshold is exceeded. However, this is not limiting. For example, it is also possible to accumulate multiple judgment results over multiple days and output an alert by observing changes in the judgment results. Specifically, for example, it is also possible to output an alert when all of the judgment results for multiple days exceed the threshold. According to this fourth modification, in addition to the effects described above, it is possible to reduce situations in which a low-stress person is mistakenly predicted as a high-stress person.

[0079] <Fourth embodiment> Next, a stress estimation device according to a fourth embodiment will be described.

[0080] The fourth embodiment is a specific example or modification of the first to third embodiments, and represents an example of calculation of a baseline, a short-term stress value, and a stress fluctuation range.

[0081] Here, the work change detection unit 1 detects the start and end times of work.

[0082] The waveform acquisition unit 2 acquires the detected biological waveform at the start of work and the detected biological waveform at the end of work. The waveform acquisition unit 2 also acquires the biological waveform during work when a certain time has passed since the timing at which the start of work was detected.

[0083] The short-term stress calculation unit 3 calculates a short-term stress value at the start of work from the biowaveform at the start of work, and calculates a short-term stress value at the end of work from the biowaveform at the end of work. The short-term stress calculation unit 3 also calculates a short-term stress value during work from the biowaveform during work.

[0084] As described above, the baseline accumulation unit 5 and the baseline calculation unit 5a each calculate a baseline based on at least the short-term stress value during work hours. Specifically, when calculating the baseline, a short-term stress value at the start of work, a short-term stress value during work hours, and a short-term stress value at the end of work hours are calculated, as shown in the bar graph in FIG. 13 . In the graph, the vertical axis represents the short-term stress value, and the horizontal axis represents time. Note that the short-term stress values ​​(10, 16, . . . , 18) in the graph are not actual measured values ​​but are expedient values ​​for easy understanding of the calculation method. Similarly, calculated values ​​such as averages and minimums based on the short-term stress values ​​in the graph are also expedient values. Furthermore, in the graph, short-term stress values ​​are calculated for a fixed period between two times. For example, a short-term stress value (10) is calculated for a fixed period T1 between two times t0 and t1. Furthermore, in the graph, two short-term stress values ​​(10, 16) are calculated for two fixed periods T1 and T2 at the start of work hours. During work hours, six short-term stress values ​​(12, 20, 24, 20, 30, 26) are calculated for six fixed periods T3, T4, ... T8. At the end of work hours, two short-term stress values ​​(20, 19) are calculated for two fixed periods T9 and T10.

[0085] Here, as shown in Figures 13 and 14, the baseline Ma may be the minimum value (13) of the average short-term stress value at the start of work (13), the average short-term stress value during work (22), and the average short-term stress value at the end of work (19).

[0086] Also, as shown in Figures 13 and 14, the baseline Mn may be the minimum value (10) of the short-term stress value at the start of work (10, 16), the short-term stress value during work (12, 20, 24, 20, 30, 26), and the short-term stress value at the end of work (20, 18).

[0087] Also, as shown in FIG. 14, the baseline Mw may be the minimum value (12) of the short-term stress values ​​(12, 20, 24, 20, 30, 26) during work.

[0088] On the other hand, the baseline removal unit 6 calculates the stress amplitude by removing the baseline from the average value of the short-term stress values ​​at least one of the start and end of work.

[0089] For example, the baseline removal unit 6 may calculate the stress amplitude by removing the baseline Mn, Ma, or Mw from the average value Sa of the short-term stress value at the start of work, as shown in FIG.

[0090] Alternatively, the baseline removal unit 6 may calculate the stress amplitude by removing two different baselines from Mn, Ma, and Mw from twice the average value of the short-term stress value at the start of work (2Sa).

[0091] The baseline removal unit 6 may calculate the stress fluctuation range by calculating the sum (Sa+Ea) of the average short-term stress value at the start of work Sa and the average short-term stress value at the end of work Ea, and removing at least the minimum short-term stress value during work from this sum as a baseline Mw. In this case, the baseline removal unit 6 may also calculate the stress fluctuation range by further removing the minimum value of the short-term stress value at the start of work, the short-term stress value during work, and the short-term stress value at the end of work as a baseline Mn.

[0092] The baseline removal unit 6 may also calculate the stress amplitude by removing two of the baselines Mn, Ma, and Mw from the sum (Sa+Ea) of the average short-term stress value Sa at the start of work and the average short-term stress value Ea at the end of work. In this case, each of the two baselines to be removed may be multiplied by a weighting coefficient (e.g., 0.2 or 1.8) that sums to 2. The baseline removal unit 6 may also calculate the stress amplitude by removing the baselines Mn, Ma, or Mw from the sum (Sa+Ea) of the average short-term stress value Sa at the start of work and the average short-term stress value Ea at the end of work.

[0093] Furthermore, the baseline removal unit 6 may calculate the stress amplitude by removing the baseline Mn, Ma, or Mw from the average value Ea of the short-term stress value at the end of work.

[0094] Alternatively, the baseline removal unit 6 may calculate the stress amplitude by removing two different baselines from Mn, Ma, and Mw from twice the average value of the short-term stress value at the end of work (2Ea).

[0095] As described above, according to the fourth embodiment, the work change detection unit 1 detects the start and end of workdays. The waveform acquisition unit 2 acquires the detected biometric waveforms at the start of workdays and the detected biometric waveforms at the end of workdays. The short-term stress calculation unit 3 calculates a short-term stress value at the start of workdays from the biometric waveforms at the start of workdays, and calculates a short-term stress value at the end of workdays from the biometric waveforms at the end of workdays. The baseline removal unit 6 calculates the sum (Sa+Ea) of the average short-term stress value at the start of workdays, Sa, and the average short-term stress value at the end of workdays, Ea, and calculates the stress fluctuation range by removing at least the minimum short-term stress value during work hours from the sum as a baseline, Mw. Therefore, in addition to the effects described above, the long-term stress level can be appropriately estimated for users who tend to experience stress at the start or end of workdays based on the short-term stress values ​​at both the start and end of workdays.

[0096] Furthermore, according to the fourth embodiment, the baseline removal unit 6 may calculate the stress fluctuation by further removing the minimum value from the short-term stress value at the start of work, the short-term stress value during work, and the short-term stress value at the end of work as a baseline Mn. In this case, in addition to the effects described above, the magnitude of the calculated stress fluctuation can be reduced.

[0097] <Fifth embodiment> FIG. 15 is a block diagram showing an example of the hardware configuration of a stress estimation device according to a fifth embodiment. The fifth embodiment is a specific example of the first to fourth embodiments, in which the stress estimation device 10 is realized by a computer. While FIG. 15 shows an example of implementation of the components corresponding to the third embodiment, the present invention is not limited to this, and components corresponding to the first, second, and fourth embodiments may also be implemented. Furthermore, in the fifth embodiment, the stress estimation device 10 is implemented in a user PC, but the present invention is not limited to this, and the stress estimation device 10 may also be implemented in a server device that can communicate with the user PC and includes a waveform acquisition unit 2.

[0098] The stress estimation device 10 includes, as hardware, a processing circuit 11, a memory 12, a display 13, a speaker 14, an input IF 15, and a communication IF 16. IF stands for interface. Each component is connected to each other via an internal bus so that they can communicate with each other.

[0099] The processing circuit 11 controls the overall operation of the stress estimation device 10. The processing circuit 11 has processors such as a CPU (Central Processing Unit), MPU (Micro Processing Unit), GPU (Graphics Processing Unit), FPU (Floating Point Unit), ASIC (Application Specific Integrated Circuit), and programmable logic device as hardware resources. For example, the processing circuit 11 executes each program loaded in the memory 12 via the processor, thereby realizing the functions of each unit corresponding to each program (e.g., short-term stress calculation unit 111, start-of-work detection unit 112, work timing determination unit 113, baseline removal unit 114, baseline calculation unit 115, stress estimation unit 116, threshold determination unit 117, and display control unit 118). The short-term stress calculation unit 111, start-of-work detection unit 112, work timing determination unit 113, baseline removal unit 114, baseline calculation unit 115, stress estimation unit 116, and threshold determination unit 117 correspond to the short-term stress calculation unit 3, start-of-work detection unit 1a, work timing determination unit 4, baseline removal unit 6, baseline calculation unit 5a, stress estimation unit 7, and threshold determination unit 9, respectively. Each unit can be realized by a processing circuit 11 consisting of a single processor, or a processing circuit 11 that combines multiple processors.

[0100] The memory 12 stores information such as data used by the processing circuit 11, a regression model dictionary 121, and programs. The regression model dictionary 121 corresponds to the regression model dictionary 8 described above. The memory 12 includes a semiconductor memory element such as a random access memory (RAM) as hardware. The memory 12 may also be a drive device that reads and writes information from and to an external storage device such as a magnetic disk (hard disk), an optical disk (CD-ROM, CD-R, DVD-ROM, DVD-R, etc.), a magneto-optical disk (MO, etc.), a semiconductor memory (USB memory, memory card, SSD), or a magnetic tape. The storage area of ​​the memory 12 may be located inside the stress estimation device 10 or in an external storage device. The memory 12 may store a baseline calculated in advance.

[0101] As described above, the program in memory 12 causes a computer to realize the functions of each unit. The program includes computer-executable instructions, and when executed by processing circuit 11, causes processing circuit 11 to perform the series of processes described with respect to each unit in FIGS. 1, 4, 6, and 11. For example, when executed by processing circuit 11, the computer-executable instructions included in the program cause processing circuit 11 to perform a stress estimation method. The stress estimation method may include steps corresponding to the functions of each unit described above. Furthermore, the stress estimation method may include the steps shown in FIGS. 2, 5, 8, and 12, as appropriate. Memory 12 is an example of a storage unit.

[0102] The program may be provided to the stress estimation device 10, which is a computer, in a state where it is stored in a computer-readable storage medium. In this case, for example, the stress estimation device 10 may further include a drive (not shown) for reading data from the storage medium and acquire the program from the storage medium. As the storage medium, for example, a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory may be used as appropriate. The storage medium may also be referred to as a non-transitory computer-readable storage medium. Alternatively, the program may be stored in a server on a communication network, and the stress estimation device 10 may download the program from the server using the communication IF 16.

[0103] The display 13 is controlled by the display control unit 118 of the processing circuit 11, and displays information such as data generated by the processing circuit 11 and data stored in the memory 12. As the display 13, for example, a liquid crystal display (LCD), a plasma display, an organic electro-luminescence display (ELD), a display of a tablet terminal, etc. can be used.

[0104] The speaker 14 outputs an alert by voice based on the alert signal output from the processing circuit 11.

[0105] The input IF 15 accepts input from a user using the stress estimation device 10, converts the accepted input into an electrical signal, and outputs the electrical signal to the processing circuit 11. The input IF 15 can be a physical operation component such as a mouse, keyboard, trackball, switch, button, joystick, touchpad, touch panel display, or microphone. The input IF 15 may also be a device that accepts input from an external input device separate from the stress estimation device 10, converts the accepted input into an electrical signal, and outputs the electrical signal to the processing circuit 11. The input IF 15 may also include a waveform acquisition unit 2 in a mouse. In this case, the waveform acquisition unit 2 may be configured to acquire a biowaveform from a user using a mouse-mounted sensor. The mouse-mounted sensor may be, for example, the photoplethysmographic sensor described above. The waveform acquisition unit 2 may be configured to acquire a biowaveform using a biosensor including at least one of an electrocardiogram sensor, a chest band sensor, a wristband sensor, and a smartphone camera (optical camera), and may be disposed in a device portable by the user. This also applies to the following modifications.

[0106] The communication IF 16 communicates various information between the stress estimation device 10 and an external device. Any communication standard can be used for this communication. The stress estimation device 10 may also transmit the estimated magnitude of long-term stress to an external device via the communication IF 16. The external device may be an operation terminal for an administrator who manages the user's long-term stress.

[0107] <Modification of the fifth embodiment> In the fifth embodiment, the functions of the stress estimation device 10 are implemented in a single computer, but this is not limiting. For example, as shown in Fig. 16, the functions of the stress estimation device 10 may be implemented in a cloud system 20. Accordingly, the stress estimation device 10 without the functions of the respective units is referred to as a user PC 10A. In other words, the stress estimation device in this modification is composed of a waveform acquisition unit 2 implemented in the mouse of the user PC 10A and a cloud system 20 having the functions of the respective units of the stress estimation device 10 described above.

[0108] The cloud system 20 is a cloud server that includes the functions of a short-term stress calculation unit 211, a start-of-work detection unit 212, a work timing determination unit 213, a baseline removal unit 214, a baseline calculation unit 215, a stress estimation unit 216, and a threshold determination unit 217. The short-term stress calculation unit 211, the start-of-work detection unit 212, the work timing determination unit 213, the baseline removal unit 214, the baseline calculation unit 215, the stress estimation unit 216, and the threshold determination unit 217 correspond to the short-term stress calculation unit 3, the start-of-work detection unit 1a, the work timing determination unit 4, the baseline removal unit 6, the baseline calculation unit 5a, the stress estimation unit 7, and the threshold determination unit 9, respectively. The cloud system 20 includes a communication interface (IF) and processing circuitry (not shown), and a memory 22 containing a regression model dictionary 221. The cloud system 20 realizes the functions of each unit by having the processing circuit execute programs stored in the memory. The cloud system 20 also communicates with the user PC 10A via the communication interface.

[0109] Accordingly, the processing circuit 11 of the user PC 10A does not include the above-mentioned units from the short-term stress calculation unit 111 to the threshold determination unit 117, and instead includes a display control unit 118 and a communication control unit 119.

[0110] The display control unit 118 has a function of controlling the display of the display 13, and for example, causes data received from the cloud system 20 to be displayed on the display 13. For example, the estimated magnitude of long-term stress, alerts, etc. can be used as the data as appropriate.

[0111] The communication control unit 119 has a function of controlling communication with the cloud system 20. For example, the communication control unit 119 transmits biological waveforms acquired by the waveform acquisition unit 2 implemented in the mouse to the cloud system 20 via the communication IF 16 during the user's start, work, and end of work hours. The communication control unit 119 also sends data received from the cloud system 20 via the communication IF to the display control unit 118.

[0112] Even with the above-described modified example, the same effects as those of the fifth embodiment can be obtained.

[0113] According to at least one of the embodiments described above, it is possible to appropriately estimate the magnitude of long-term stress for a user who is prone to experiencing stress at the start or end of work. This also applies to at least one of the modified examples described above.

[0114] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, as well as within the scope of the invention described in the claims and their equivalents. [Explanation of symbols]

[0115] 10...Stress estimation device, 1...Work change detection unit, 1a,112,212...Start of work detection unit, 2...Waveform acquisition unit, 3,111,211...Short-term stress calculation unit, 4,113,213...Work timing determination unit, 5...Baseline accumulation unit, 5a,115,215...Baseline calculation unit, 6,114,214...Baseline removal unit, 7,116,216...Stress estimation unit, 8,121,221...Regression model dictionary, 9,117,217...Threshold determination unit, 11...Processing circuit, 12,22...Memory, 13...Display, 14...Speaker, 15...Input IF, 16...Communication IF, 118...Display control unit, 119...Communication control unit.

Claims

1. a detection unit configured to detect at least one of a start time and an end time of a workday for a user who works during a workday between a start time and an end time of a workday; an acquisition unit that acquires a biological waveform of the user based on the detected result; a calculation unit that calculates a short-term stress value from the biological waveform; a baseline calculation unit that calculates a baseline based on at least the short-term stress value during work; a removal unit that calculates the stress fluctuation range of the user by removing the baseline from the calculated short-term stress value; an estimation unit that estimates the magnitude of the long-term stress of the user based on the stress fluctuation range; A stress estimation device comprising:

2. The detection unit detects the start of work, The acquisition unit acquires the detected biological waveform at the start of work, the calculation unit calculates a short-term stress value at the start of work from the biological waveform at the start of work, The stress estimation device according to claim 1 , wherein the removal unit calculates the stress fluctuation range at the start of work by removing the baseline from the short-term stress value at the start of work.

3. The detection unit detects the end of work, The acquisition unit acquires the biological waveform at the detected end of work, the calculation unit calculates a short-term stress value at the end of work from the biological waveform at the end of work, The stress estimation device according to claim 1 , wherein the removal unit calculates the stress fluctuation range at the end of work by removing the baseline from the short-term stress value at the end of work.

4. The stress estimation device according to claim 1 , wherein the baseline is the minimum value among the short-term stress value at the start of work, the short-term stress value during work, and the short-term stress value at the end of work.

5. 4. The stress estimation device according to claim 1, wherein the baseline is the minimum value among the average value of the short-term stress value at the start of work, the average value of the short-term stress value during work, and the average value of the short-term stress value at the end of work.

6. The detection unit detects the start time and the end time of work, the acquisition unit acquires the detected biological waveform at the start of work and the detected biological waveform at the end of work, the calculation unit calculates a short-term stress value at the start of work from the biological waveform at the start of work, and calculates a short-term stress value at the end of work from the biological waveform at the end of work, 2. The stress estimation device of claim 1, wherein the elimination unit calculates the sum of the average short-term stress value at the start of work and the average short-term stress value at the end of work, and calculates the stress fluctuation range by removing at least the minimum value of the short-term stress value during work from the sum as the baseline.

7. 7. The stress estimation device according to claim 6, wherein the elimination unit calculates the stress fluctuation range by further eliminating the minimum value among the short-term stress value at the start of work, the short-term stress value during work, and the short-term stress value at the end of work as the baseline.

8. The stress estimation device according to claim 1 , wherein the detection unit detects at least one of the start time and the end time of work in response to an operation of the user on an office device used for work.

9. The stress estimation device according to claim 8 , wherein the detection unit detects the start of workday based on a start-up operation by the user, and detects the end of workday based on an end-of-work operation by the user.

10. The stress estimation device according to claim 9 , wherein the detection unit, when detecting the start time of work, distinguishes between a morning start time of work and an afternoon start time of work based on the activation operation and the time of day.

11. The stress estimation device according to claim 1 , further comprising a sending unit that sends the calculated short-term stress value to at least one of the baseline calculation unit and the removal unit based on the detection result.

12. a storage unit that stores in advance a regression model that represents the relationship between stress fluctuation and the magnitude of long-term stress; Further provided with The stress estimation device according to claim 11 , wherein the estimation unit estimates the magnitude of the long-term stress of the user based on the regression model.

13. The storage unit stores the regression model in advance for each type of stress fluctuation range, The stress estimation device according to claim 12 , wherein the estimation unit estimates the magnitude of the long-term stress of the user based on the regression model according to the type of the calculated stress fluctuation range.

14. an output unit that outputs an alert when the magnitude of the estimated long-term stress exceeds a threshold; The stress estimation device according to claim 11, further comprising:

15. the acquisition unit acquires the biological waveform from the user using a biological sensor; The stress estimation device according to claim 11 , wherein the biosensor includes at least one of an electrocardiogram sensor, a photoplethysmogram sensor, a chest band sensor, a wrist band sensor, a mouse-mounted sensor, and an optical camera.

16. the acquisition unit is disposed in an office device used for work or a device portable to the user, The stress estimation device according to claim 11 , wherein the calculation unit, the detection unit, the transmission unit, the baseline calculation unit, the removal unit, and the estimation unit are arranged on a cloud server.

17. a detection unit detecting at least one of a start time and an end time of a workday for a user who works during a workday between a start time and an end time of a workday; an acquisition unit acquiring a biological waveform of the user based on the detected result; A calculation unit calculates a short-term stress value from the biological waveform; a baseline calculation unit calculating a baseline based on at least the short-term stress value during work; a removal unit removing the baseline from the calculated short-term stress value to calculate a stress fluctuation range of the user; an estimation unit estimating a magnitude of long-term stress of the user based on the stress fluctuation; A stress estimation method comprising:

18. a function of detecting at least one of a start time and an end time for a user who works between the start time and the end time of work; a function of acquiring a biological waveform of the user from a sensor based on the detected result; A function of calculating a short-term stress value from the biological waveform; A function of calculating a baseline based on at least the short-term stress value during work; a function of calculating a stress fluctuation range of the user by removing the baseline from the calculated short-term stress value; a function of estimating the magnitude of the user's long-term stress based on the stress fluctuation; A program to make the above happen on a computer.

Images