Cognitive function estimation device and cognitive function estimation program

Abstract

To provide a cognitive function estimation device and a cognitive function estimation program capable of estimating a cognitive function state of a user with a high degree of precision.SOLUTION: A cognitive function estimation device 40A includes an emotion detection part 41c, an action detection part 41f, and a long term analysis part 41h as an estimation part. The emotion detection part 41c detects aggravation of an emotion of a user Y on the basis of biological information on the user Y. The action detection part 41f detects a related action, which is related to the aggravation of the emotion at a time point later than the time when the aggravation of the emotion of the user Y is detected. The long term analysis part 41h estimates deterioration in the cognitive function of the user Y on the basis of the result of the detection by the emotion detection part 41c and the result of the detection by the action detection part 41f.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a cognitive function estimation device and a cognitive function estimation program. [Background technology]

[0002] In recent years, as the aging of the population progresses, dealing with dementia has become a major social issue. In order to provide appropriate care for dementia patients, it is necessary to understand the state of decline in cognitive function of each patient.

[0003] In view of this, there is a technology for estimating the cognitive function level of a subject, whose cognitive function decline is to be grasped, based on the emotional level, voice data, gaze, and other behavioral details of the subject (for example, Patent Document 1). By using this technology, the cognitive function level can be estimated from various information related to the subject. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-58231 Summary of the Invention [Problem to be solved by the invention]

[0005] When the emotional level of dementia patients deteriorates due to a decline in cognitive function, they may shout, throw things, and do other things that they would not do when they are emotionally stable. In other words, the state of cognitive function may affect the subject's emotional level and the behavior that results from it.

[0006] However, the above-mentioned technology has a problem in that it is not possible to estimate the state of decline in cognitive function by taking into account the relationship between the deterioration of emotional level and the content of behavior.

[0007] The present invention aims to provide a cognitive function estimation device and a cognitive function estimation program capable of accurately estimating the state of a user's cognitive function. [Means for solving the problem]

[0008] In order to achieve the above-mentioned object, the cognitive function estimation device of the present invention comprises an emotion detection unit that detects a deterioration in a user's emotions based on the user's biometric information, a behavior detection unit that detects associated behaviors related to the deterioration in the user's emotions at a time after the deterioration in the user's emotions is detected, and an estimation unit that estimates a decline in cognitive function of the user based on the detection results by the emotion detection unit and the detection results by the behavior detection unit.

[0009] In addition, the cognitive function estimation program of the present invention enables a cognitive function estimation device to realize an emotion detection function that detects a deterioration in the user's emotions based on the user's biometric information, a behavior detection function that detects associated behaviors related to the deterioration in the user's emotions at a time after the deterioration in the user's emotions is detected, and an estimation function that estimates a decline in the user's cognitive function based on the detection results by the emotion detection function and the detection results by the behavior detection function. Effect of the Invention

[0010] According to the cognitive function estimation device and cognitive function estimation program of the present invention, the state of a user's cognitive function can be estimated with high accuracy. [Brief description of the drawings]

[0011] [Figure 1] 1 is an overall diagram showing the configuration of a cognitive function estimation system using a cognitive function estimation device according to a first embodiment of the present invention. [Diagram 2] 4 is a flowchart showing a cognitive function estimation process executed by the cognitive function estimation device according to the first and second embodiments of the present invention. [Diagram 3] 4 is a flowchart showing a short-term analysis process executed by the cognitive function estimation device according to the first embodiment of the present invention. [Figure 4]FIG. 11 is an overall diagram showing the configuration of a cognitive function estimation system using a cognitive function estimation device according to a second embodiment of the present invention. [Diagram 5] 13 is a flowchart showing a short-term analysis process executed by a cognitive function estimation device according to a second embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, a cognitive function estimation system using a cognitive function estimation device according to an embodiment of the present invention will be described.

[0013] First Embodiment Configuration of the cognitive function estimation system according to the first embodiment 1 is a block diagram showing the configuration of a cognitive function estimation system 1A according to this embodiment. The cognitive function estimation system 1A includes a plurality of biological information measuring sensors 10, a plurality of behavioral information measuring sensors 20, and a cognitive function estimation device 40A. The cognitive function estimation system 1A is configured, for example, with a portable information terminal such as a personal computer (PC) or a smartphone, a server, and the like. The plurality of biological information measuring sensors 10 and the plurality of behavioral information measuring sensors 20 are, for example, sensors wirelessly connected to the cognitive function estimation system 1A, and as an example, are included in a smart watch worn by a user Y.

[0014] Each of the plurality of biometric sensors 10 measures biometric information related to the deterioration of the emotion of the user Y, who is the subject of cognitive function estimation. Here, the deterioration of emotion indicates that the emotion of the user Y at a certain point in time changes to an emotion that is generally considered to be bad at a later point in time, for example, that the emotion of the user Y changes from a calm emotion to an irritated emotion. The biometric sensor 10 is composed of, for example, a pulse meter 10a, a blood pressure meter 10b, a skin potential meter 10c, an autonomic nerve meter 10d that measures the autonomic nerve activity level, and an emotion recognition device 10e, which are attached to the body of the user Y. The biometric sensor 10 is not limited to a device or apparatus attached to the body of the user Y, and may be any device that measures biometric information related to the deterioration of the emotion of the user Y. The emotion recognition device 10e photographs the face of the user Y, reads the facial expression, and indicates the good or bad emotion in a numerical value. Hereinafter, when it is not specified which of these devices it is, it will be described as the biometric sensor 10.

[0015] The behavioral information measuring sensors 20 measure behavioral information for detecting related behaviors of the user Y. Here, the related behaviors refer to behaviors related to the deterioration of emotions, for example, behaviors that the user Y could not bear to endure after the deterioration of his / her emotions. The behavioral information measuring sensors 20 are, for example, composed of a sound collecting device 20a installed near the user Y, or an acceleration sensor 20b installed on the body of the user Y. The acceleration sensor 20b is, for example, installed on the arm of the user Y. Hereinafter, when it is not specified which of these devices or sensors it is, it will be described as the behavioral information measuring sensor 20.

[0016] The cognitive function estimation device 40A includes a CPU (Central Processing Unit) 41A, a storage unit 42A, and a display unit 43. The storage unit 42A is configured with, for example, a RAM (Random Access Memory) or a ROM (Read Only Memory). The CPU 41A includes, as functional blocks realized by executing a program, a biological information acquisition unit 41a, an emotion threshold calculation unit 41b, an emotion detection unit 41c, a behavior information acquisition unit 41d, a behavior threshold calculation unit 41e, a behavior detection unit 41f, a short-term analysis unit 41g and a long-term analysis unit 41h as an estimation unit, and a display information generation unit 41i.

[0017] The memory unit 42A has a biological information memory unit 42a, an emotion threshold memory unit 42b, an emotion detection information memory unit 42c, a behavioral information memory unit 42d, a behavioral threshold memory unit 42e, a related behavioral information memory unit 42f, a short-term analysis information memory unit 42g, and a long-term analysis information memory unit 42h.

[0018] The biological information acquiring unit 41a acquires biological information of the user Y measured by the biological information measuring sensor 10. The biological information acquiring unit 41a stores the acquired information in the biological information storage unit 42a.

[0019] The emotion threshold calculation unit 41b calculates an emotion threshold, which is a threshold for determining whether the emotion of the user Y has deteriorated, based on a plurality of pieces of biometric information within a predetermined period stored in the biometric information storage unit 42a. The calculation of the emotion threshold will be described in detail later. The emotion threshold calculation unit 41b stores the calculated emotion threshold in the emotion threshold storage unit 42b.

[0020] The emotion detection unit 41c detects the deterioration of the emotion of the user Y based on the biometric information of the user Y acquired by the biometric information acquisition unit 41a. Specifically, when the biometric information acquired by the biometric information acquisition unit 41a exceeds the emotion threshold stored in the emotion threshold storage unit 42b during the cognitive function estimation process described later, the emotion detection unit 41c detects that the emotion of the user Y has deteriorated. Here, the method in which the emotion detection unit 41c detects the deterioration of the emotion of the user Y is not limited to the use of the emotion threshold. For example, the emotion detection unit 41c may detect the deterioration of the emotion of the user Y by using an emotion analysis model (not shown) that has been learned in advance. In this case, for example, the deterioration of the emotion of the user Y is detected by inputting the biometric information of the user Y into the emotion analysis model. In this case, the cognitive function estimation device 40A may not include the emotion threshold calculation unit 41b and the emotion threshold storage unit 42b. The emotion detection unit 41c stores the detected information in the emotion detection information storage unit 42c.

[0021] The behavioral information acquiring unit 41d acquires behavioral information of the user Y measured by the behavioral information measuring sensor 20 at a time after the deterioration of the emotion of the user Y is detected. The timing at which the behavioral information acquiring unit 41d acquires behavioral information is not limited to this, and for example, the behavioral information of the user Y measured by the behavioral information measuring sensor 20 may be acquired continuously. The behavioral information acquiring unit 41d stores the acquired information in the behavioral information storage unit 42d.

[0022] The behavior threshold calculation unit 41e calculates a behavior threshold, which is a threshold for determining whether or not the user Y has performed a related behavior related to the deterioration of emotions, based on the behavior information for a predetermined period stored in the behavior information storage unit 42d. The calculation of the behavior threshold will be described in detail later. The behavior threshold calculation unit 41e stores the calculated behavior threshold in the behavior threshold storage unit 42e.

[0023] The behavior detection unit 41f detects a related behavior related to the deterioration of the user Y's emotion based on the behavior information of the user Y acquired by the behavior information acquisition unit 41d at a time point after the deterioration of the user Y's emotion is detected. Specifically, when the behavior information acquired by the behavior information acquisition unit 41d exceeds the behavior threshold stored in the behavior threshold storage unit 42e during the cognitive function estimation process described later, the behavior detection unit 41f detects that the user Y has performed the related behavior. Here, the method in which the behavior detection unit 41f detects the related behavior of the user Y is not limited to the use of the behavior threshold. For example, the behavior detection unit 41f may detect the related behavior of the user Y by using a behavior analysis model (not shown) that has been learned in advance. In this case, for example, the behavior information of the user Y is input into the behavior analysis model to detect that the user Y has performed the related behavior. In this case, the cognitive function estimation device 40A may not include the behavior threshold calculation unit 41e and the behavior threshold storage unit 42e. The behavior detection unit 41f stores the detected information in the related behavior information storage unit 42f.

[0024] The short-term analysis unit 41g generates short-term analysis information, which is short-term analysis information related to the cognitive function of the user Y, based on the information stored in the emotion detection information storage unit 42c and the information stored in the related behavior information storage unit 42f. The short-term analysis unit 41g stores the generated information in the short-term analysis information storage unit 42g.

[0025] The long-term analysis unit 41h generates long-term analysis information, which is long-term analysis information regarding the cognitive function of the user Y, based on the information stored in the short-term analysis information storage unit 42g. The long-term analysis unit 41h stores the generated information in the long-term analysis information storage unit 42h.

[0026] The display information generating unit 41i generates display information for displaying the analysis information generated by the short-term analysis unit 41g and the long-term analysis unit 41h.

[0027] The display unit 43 displays the display information generated by the display information generation unit 41i. However, the cognitive function estimation device 40A does not need to include the display unit 43, and in that case, the information generated by the display information generation unit 41i, the information stored in the short-term analysis information storage unit 42g, or the information stored in the long-term analysis information storage unit 42h may be output to another device (not shown).

[0028] <Operation of the cognitive function estimation system according to one embodiment> The cognitive function estimation device 40A first performs a process of setting an emotion threshold and a behavior threshold as preprocessing of the cognitive function estimation process.

[0029] When the cognitive function estimation device 40A sets an emotion threshold, first, the biological information acquisition unit 41a sequentially acquires biological information of the user Y measured by the biological information measurement sensor 10 at predetermined time intervals.

[0030] In this embodiment, the items of bioinformation acquired by the bioinformation acquiring unit 41a are the pulse measured by the pulse meter 10a, which is the bioinformation measuring sensor 10, the blood pressure measured by the blood pressure measuring instrument 10b, the skin potential measured by the skin potential measuring instrument 10c, the autonomic nerve activity measured by the autonomic nerve measuring instrument 10d, and a value indicating good or bad emotion based on facial expression shown by the emotion recognition device 10e.

[0031] The time interval for acquiring these pieces of biometric information may be different for each item, for example, every second for pulse rate and values ​​indicating good or bad emotions based on facial expressions, and every 10 minutes for blood pressure, skin potential, and autonomic nerve activity. The biometric information acquiring unit 41a adds the respective measurement time information to the acquired measurement value of each item and stores them in the biometric information storage unit 42a.

[0032] When the emotion threshold calculation unit 41b recognizes that a sufficient amount of biometric information for calculating the emotion threshold, for example one month's worth of biometric information, has been stored in the biometric information storage unit 42a, it uses the stored biometric information to calculate an emotion threshold, which is a threshold for determining whether or not user Y's emotion has deteriorated.

[0033] The emotion threshold calculation unit 41b acquires the biometric information of the user Y for each item for a predetermined time period in a day, for example, every hour for 24 hours, from among the items stored in the biometric information storage unit 42a. Here, the predetermined time period in a day may be a morning time period, a daytime time period, or a nighttime time period. The emotion threshold calculation unit 41b acquires a predetermined percentage of the pulse, blood pressure, and autonomic nerve activity from the lower end, for example, 2 / 3 of the entire time period, a predetermined percentage of the skin potential from the higher end, for example, 2 / 3 of the entire time period, and a predetermined percentage of the value indicating the good or bad emotion based on facial expression from the better end, for example, 2 / 3 of the entire time period, from among the biometric information whose measurement time corresponds to each time period, as biometric information of the user Y in a calm emotional state.

[0034] The emotion threshold calculation unit 41b calculates the average value and standard deviation of the measurement value of each item acquired for each time period, and calculates the upper limit value of the state in which the user Y's emotions are calm for each time period and each item based on these. For the autonomic nerve activity level, the emotion threshold is set to the value at which the sympathetic nerve transitions to a dominant state. The emotion threshold calculation unit 41b stores the emotion threshold calculated for each time period and each item in the emotion threshold storage unit 42b. By calculating the emotion threshold for each time period in this way, it is possible to set the threshold taking into account the fluctuation of emotions during the day. However, the emotion threshold calculation unit 41b may calculate the emotion threshold for each item from the bioinformation for all items stored in the bioinformation storage unit 42a without considering the time period.

[0035] In addition, when the cognitive function estimation device 40A sets a behavioral threshold, the behavioral information acquisition unit 41d first sequentially acquires behavioral information of user Y measured by the behavioral information measurement sensor 20 at predetermined time intervals, for example, every second, and stores the information in the behavioral information storage unit 42d after adding measurement time information to each piece of information.

[0036] In this embodiment, the items of behavioral information acquired by the behavioral information acquisition unit 41d are voice information of user Y measured by the sound collection device 20a, which is the behavioral information measurement sensor 20, and arm movement information of user Y measured by the acceleration sensor 20b.

[0037] When the behavior threshold calculation unit 41e recognizes that a sufficient amount of behavior information, for example, one month's worth of behavior information, has been stored in the behavior information storage unit 42d to calculate a behavior threshold, the behavior threshold calculation unit 41e uses the stored behavior information to calculate a behavior threshold, which is a threshold for determining whether or not user Y has engaged in related behavior related to a deterioration in emotions.

[0038] The behavior threshold calculation unit 41e acquires behavior information of each item of the user Y's emotionally calm state for each predetermined time period, for example, every hour, from among the behavior information stored in the behavior information storage unit 42d. As behavior information of the user Y's calm state, the behavior threshold calculation unit 41e acquires a predetermined percentage of the measurement values ​​from the lower voice volume, for example, 2 / 3 of the entire time period, from the behavior information whose measurement time corresponds to each time period, and acquires a predetermined percentage of the measurement values ​​from the slowest arm movement speed, for example, 2 / 3 of the entire time period, from the behavior information whose measurement time corresponds to each time period.

[0039] The behavior threshold calculation unit 41e calculates the average value and standard deviation of the acquired measurement value of each item for each time period, and calculates the upper limit value of the state in which the user Y's emotions are calm for each time period and each item based on the average value and standard deviation. The behavior threshold calculation unit 41e stores the calculated behavior threshold for each time period and each item in the behavior threshold storage unit 42e. This completes the pre-processing of the cognitive function estimation process.

[0040] 2 is a flowchart showing the cognitive function estimation process executed by the cognitive function estimation device 40 A. When starting the cognitive function estimation process, the cognitive function estimation device 40 A first executes a short-term analysis process (step S1).

[0041] 3 is a flowchart showing a short-term analysis process executed by the cognitive function estimation device 40 A. When the short-term analysis process starts, the biological information acquisition unit 41 a sequentially acquires biological information of the user Y measured by the biological information measurement sensor 10 at predetermined time intervals (step S11).

[0042] The items of bioinformation acquired by the bioinformation acquisition unit 41a are the pulse measured by the pulse meter 10a, which is the bioinformation measurement sensor 10, the blood pressure measured by the blood pressure meter 10b, the skin potential measured by the skin potential meter 10c, the autonomic nerve activity measured by the autonomic nerve meter 10d, and a value indicating good or bad emotion based on facial expression shown by the emotion recognition device 10e.

[0043] The time interval for acquiring these pieces of biometric information may be different for each item, for example, every second for pulse rate and values ​​indicating good or bad emotions based on facial expressions, and every 10 minutes for blood pressure, skin potential, and autonomic nerve activity. The biometric information acquiring unit 41a adds the respective measurement time information to the acquired measurement value of each item.

[0044] Each time the biometric information acquisition unit 41a acquires a new measurement value for any item, the emotion detection unit 41c judges whether or not each measurement value for each item of biometric information exceeds the emotion threshold for the corresponding time period and item stored in the emotion threshold storage unit 42b (step S12). For example, when the execution time of step S11 is 13:30, the emotion detection unit 41c judges whether or not each measurement value for each item of biometric information exceeds the emotion threshold for each item for the 13:00 time period stored in the emotion threshold storage unit 42b.

[0045] If a predetermined number or more of the latest measurement values ​​of each item, for example, two or more items, exceed the corresponding emotion threshold ("YES" in step S12), the emotion detection unit 41c determines that the user Y is, for example, irritated, that is, the emotion of the user Y is deteriorating (step S13). If a predetermined number or more of the measurement values ​​of each item do not exceed the corresponding emotion threshold ("NO" in step S12), the process returns to step S11.

[0046] When the emotion detection unit 41c detects the deterioration of the emotion of the user Y, the behavior information acquisition unit 41d starts acquiring behavior information of the user Y measured by the behavior information measurement sensor 20 (step S14). In addition, when the behavior information acquisition unit 41d constantly acquires behavior information, the process of step S14 may be omitted and the process may proceed to step S15. The behavior information acquisition unit 41d acquires behavior information of the user Y measured by the behavior information measurement sensor 20 at predetermined time intervals, for example, every second. The items of behavior information acquired by the behavior information acquisition unit 41d are voice information of the user Y measured by the sound collection device 20a, which is the behavior information measurement sensor 20, and arm movement information of the user Y measured by the acceleration sensor 20b. The behavior information acquisition unit 41d adds each measurement time information to the acquired behavior information of each item.

[0047] Each time the behavior information acquisition unit 41d acquires behavior information, the behavior detection unit 41f determines, for each item of the acquired behavior information, whether or not each measurement value exceeds the behavior threshold for the corresponding time period and item stored in the behavior information storage unit 42d (step S15).

[0048] When one or more of the measured values ​​of the acquired items exceed the corresponding behavior thresholds ("YES" in step S15), the behavior detection unit 41f determines that the user Y cannot suppress the worsening of emotions and has performed a related behavior related to the worsening of emotions (step S16). For example, when it is determined that the user Y has spoken louder than usual based on the voice information of the user Y measured by the sound collection device 20a, the behavior detection unit 41f determines that the user Y has performed a related behavior. Alternatively, when it is determined that the user Y has made a large arm movement based on the movement information of the user Y measured by the acceleration sensor 20b, the behavior detection unit 41f determines that the user Y has performed a related behavior.

[0049] At this time, the behavior detection unit 41f may acquire a value measured by the emotion recognition device 10e, which is one of the bio-information measuring sensors 10, and use the value to assist in determining whether the user Y could not bear the worsening emotion and took the related behavior. For example, if the emotion recognition device 10e recognizes that the user Y has an angry expression when the user Y speaks loudly or moves his / her arms widely, the behavior detection unit 41f may determine that the user Y could not bear the worsening emotion and took the related behavior.

[0050] The processes of steps S15 and S16 are repeated until the biological information of each item falls below the emotion threshold ("NO" in step S17). When the biological information of each item falls below the emotion threshold ("YES" in step S17), the cognitive function estimation device 40A returns to step S11 and repeats the processes of steps S11 to S16. When the cognitive function estimation device 40A detects in step S17 that the number of items whose measured values ​​exceed the emotion threshold falls below the number used in the determination in step S12, the cognitive function estimation device 40A may return to step S11. Alternatively, the cognitive function estimation device 40A may return to step S11 when a predetermined period has elapsed since it was determined in step S13 that the emotion of the user Y has deteriorated.

[0051] Returning to the flowchart of Fig. 2, the short-term analysis unit 41g generates short-term analysis information, which is short-term analysis information on the cognitive function of the user Y, based on the information stored in the emotion detection information storage unit 42c and the information stored in the related behavior information storage unit 42f (step S2). This "short period" is a first period T1, which is, for example, "one day", but is not limited to this and may be one hour or one month.

[0052] In the short-term analysis process, the short-term analysis unit 41g calculates the values ​​of five items as short-term analysis information in the first period T1: (1) ratio of time during which the mood worsens, (2) duration of time during which the mood worsens, (3) number of occurrences of mood worsening, (4) number of occurrences of related behavior, and (5) time from when the mood worsens to when the related behavior occurs. The specific calculation contents of each item are described below.

[0053] (1) Percentage of time spent feeling aggravated The ratio of time during which the user's emotions are determined to be worsened is the ratio of time during which the user's emotions are determined to be worsened to the 24 hours of a day, which is the target period. This "time during which the user's emotions are determined to be worsened" is the time from the time when the user's emotions are first determined to be worsened in step S13 during the repetition of the processes of steps S11 to S16 described above to the time when the user's emotions are determined to be no worse thereafter, and is calculated in seconds. The "time when the user's emotions are first determined to be worsened" and the "time when the user's emotions are determined to be no worsened" are measurement time information added to the measurement values ​​of the items that triggered these determination processes.

[0054] (2) Duration of emotional deterioration The duration of worsening emotions is the average value of the aforementioned "time during which emotions were judged to be worsening" that occurred during the target day, and is calculated in seconds.

[0055] (3) Number of times emotions worsened The number of times of worsening mood is the number of times "times when it is determined that the mood is worsening" that occurred during the target period, which is one day. The number of times of worsening mood is counted as one time during the repetition of the processes of steps S11 to S16 described above, from when it is first determined that the mood of user Y is worsening in step S13 until it is determined that the mood of user Y is not worsening thereafter.

[0056] (4) Number of related actions The number of related behaviors is the number of related behaviors that occurred during a target period of one day. The number of related behaviors is counted as one occurrence during the repetition of the above-described steps S11 to S16, from when it is first determined in step S16 that user Y has performed a related behavior until it is determined that user Y has not performed any related behavior thereafter.

[0057] (5) Time from emotional deterioration to related behavior The time from the worsening of emotion to the related behavior is the average or total time from when the user Y's emotion worsens to when the related behavior occurs during a day, which is the target period, and is calculated in seconds. This time from the worsening of emotion to the related behavior may be calculated during the determination process of step S16 described above.

[0058] The short-term analysis unit 41g stores the values ​​of the above items (1) to (5) calculated in the short-term analysis process in the short-term analysis information storage unit 42g as short-term analysis information. Note that the items calculated in the short-term analysis process are not limited to the above items (1) to (5). For example, the short-term analysis unit 41g may calculate the related behavior duration as the short-term analysis information, which is the time from when it is first determined that the user Y has performed the related behavior to when it is determined that the user Y has not performed the related behavior thereafter.

[0059] Next, the long-term analysis unit 41h generates long-term analysis information, which is long-term analysis information regarding the cognitive function of the user Y, based on the information stored in the short-term analysis information storage unit 42g (step S3).

[0060] In the long-term analysis process, the long-term analysis unit 41h calculates, as long-term analysis information, (a) the values ​​of the six items, namely, the above items (1) to (5) and (6) the occurrence rate of related behavior in the second period T2, and (b) cognitive function estimation information in the third period T3. The second period T2 is a period longer than the first period T1, for example, one month, and the third period T3 is a period even longer than the second period T2, for example, one year.

[0061] The above (6) related behavior occurrence rate is the ratio of the total number of times related behavior occurred within a specified period after a deterioration in emotions to the total number of times emotions worsened during the second period T2, which is one month.

[0062] The long-term analysis unit 41h may execute the process of calculating the values ​​of the six items (1) to (6) in the second period T2 of (a) on a daily basis or on a monthly basis. When the long-term analysis unit 41h executes the process of (a) on a daily basis, the long-term analysis unit 41h executes the long-term analysis process by shifting the target period by one day, for example, on February 1st, the target period is one month from January 1st to January 31st, on February 2nd, the target period is one month from January 2nd to February 1st, and on February 3rd, the target period is one month from January 3rd to February 2nd.

[0063] Furthermore, when the long-term analysis unit 41h executes the process (a) every month, it executes the long-term analysis process by shifting the target period by one month, for example, on February 1st it sets the target period to one month from January 1st to January 31st, and on March 1st it sets the target period to February 1st to February 28th.

[0064] The cognitive function estimation information in the third period T3 of (b) above will be described. The long-term analysis unit 41h obtains a regression line from information for one year, which is the third period T3, for the values ​​of each of the items (1) to (6) for each second period T2 calculated in the process of (a).

[0065] The long-term analysis unit 41h sets a flag for items where the regression line is inclined in a direction of increasing with time for item (1) percentage of time of worsening emotion, item (2) duration of worsening emotion, item (3) number of times of worsening emotion, item (4) number of times of related behavior, and item (6) percentage of occurrence of related behavior. The long-term analysis unit 41h also sets a flag for items where the regression line is inclined in a direction of decreasing with time for item (5) time from worsening emotion to related behavior.

[0066] (1) The fact that the regression line for the proportion of time spent with worsening emotions is trending in an increasing direction indicates that the time during which user Y's emotions worsen is likely to become longer. (2) The fact that the regression line for the duration of time spent with worsening emotions is trending in an increasing direction indicates that the time from when user Y's emotions worsen to when they are resolved is likely to become longer. (3) The fact that the regression line for the number of times emotions worsen is trending in an increasing direction indicates that the number of times user Y's emotions worsen is likely to become longer. (4) The fact that the regression line for the number of related behaviors is trending in an increasing direction indicates that the number of times user Y was unable to tolerate his / her emotions when they worsened is likely to become longer. (5) The fact that the regression line for the time from when emotions worsen to when related behaviors are trending in a decreasing direction indicates that the time from when user Y's emotions worsen to when he / she can no longer tolerate them is likely to become shorter. (6) The fact that the regression line for the proportion of occurrences of related behaviors is trending in an increasing direction indicates that the proportion of times user Y was unable to tolerate his / her emotions when they worsened is likely to become longer.

[0067] These tendencies can be considered to be due to a decline in cognitive function, so the long-term analysis unit 41h generates cognitive function estimation information that estimates that the greater the number of flags, the more advanced the decline in cognitive function. The long-term analysis unit 41h may estimate the decline in cognitive function based on other estimation methods. The long-term analysis unit 41h as an estimation unit estimates the decline in cognitive function based on at least one of (1) the proportion of time during which emotions worsen, (2) the duration of emotions worsening, (3) the number of times emotions worsen, (4) the number of related behaviors, (5) the time from emotions worsening to related behaviors, and (6) the proportion of related behaviors occurring. In other words, the long-term analysis unit 41h as an estimation unit estimates the decline in cognitive function based on the detection results by the emotion detection unit 41c and the detection results by the behavior detection unit 41f.

[0068] In addition, since people with dementia generally tend to become less tolerant of dementia-related behaviors as their symptoms progress, the long-term analysis unit 41h may generate cognitive function estimation information by attaching a higher importance to (6) the change in the rate of occurrence of dementia-related behaviors among the above items (1) to (6) and using the changes in the other items (1) to (5) as auxiliary measures.

[0069] Moreover, the long-term analysis unit 41h may set the third period T3 even longer and analyze the measured values ​​of the items (1) to (6) over several years to generate cognitive function estimation information.

[0070] The long-term analysis unit 41h stores, in the long-term analysis information memory unit 42h, as long-term analysis information, (a) the values ​​of the six items, namely, the above items (1) to (5) in the second period T2 and (6) the occurrence rate of related behaviors, calculated in the long-term analysis processing, and (b) cognitive function estimation information in the third period T3.

[0071] The display information generation unit 41i generates display information for displaying the information stored in at least one of the short-term analysis information storage unit 42g or the long-term analysis information storage unit 42h based on instructions from a monitor or the like monitoring user Y, and displays the information on the display unit 43 (step S4).

[0072] According to the first embodiment described above, the state of the user's cognitive function can be accurately estimated by monitoring the user's emotional deterioration and the occurrence state of related behaviors related to the emotional deterioration in the short term and over the long term.

[0073] Second Embodiment Configuration of the cognitive function estimation system according to the second embodiment 4 is a block diagram showing the configuration of a cognitive function estimation system 1B according to this embodiment. The cognitive function estimation system 1B has a cognitive function estimation device 40B instead of the cognitive function estimation device 40A, and further has an environmental information acquisition sensor 30. Other than this, the cognitive function estimation system 1B has the same configuration as the cognitive function estimation system 1A described in the first embodiment, and therefore detailed descriptions of components having the same functions will be omitted.

[0074] The environmental information acquisition sensor 30 is configured, for example, by a sound collection device installed near the user Y, and acquires information about the surroundings of the user Y. The configuration of the environmental information acquisition sensor 30 is not limited to a sound collection device, and may be configured, for example, by an imaging device. When the environmental information acquisition sensor 30 is a sound collection device, it may be configured as the same device as the sound collection device 20a, which is one of the behavioral information measurement sensors 20.

[0075] The CPU 41B of the cognitive function estimation device 40B has an environmental information acquisition unit 41j and a factor detection unit 41k in addition to the configuration of the CPU 41A of the cognitive function estimation device 40A described in the first embodiment. The storage unit 42B has a factor information storage unit 42i in addition to the configuration of the storage unit 42A.

[0076] The environmental information acquisition unit 41j acquires information on the surrounding environment of the user Y measured by the environmental information acquisition sensor 30. The factor detection unit 41k detects factors in the surrounding environment of the user Y that may worsen the emotions of the user Y, based on the information acquired by the environmental information acquisition unit 41j.

[0077] The factor information storage unit 42i stores the detection result by the factor detection unit 41k, specifically, information on the factors that worsen the emotion detected by the factor detection unit 41k.

[0078] In this embodiment, the short-term analysis unit 41g generates short-term analysis information, which is short-term analysis information regarding the cognitive function of the user Y, based on the information stored in the emotion detection information storage unit 42c, the information stored in the related behavior information storage unit 42f, and the information stored in the factor information storage unit 42i. That is, the long-term analysis unit 41h as an estimation unit estimates a decline in cognitive function based on the detection result by the emotion detection unit 41c, the detection result by the behavior detection unit 41f, and the detection result by the factor detection unit 41k.

[0079] <Operation of the cognitive function estimation system according to the second embodiment> Fig. 5 is a flowchart showing the short-term analysis process executed by the cognitive function estimation device 40 B. In Fig. 5, the process of steps S21 to S25 is similar to the process of steps S11 to S15 described in the first embodiment, and therefore detailed description thereof will be omitted.

[0080] In this embodiment, when acquisition of the behavioral information of the user Y is started in step S24, the environment information acquisition unit 41j also starts a process of acquiring information on the surrounding environment of the user Y measured by the environment information acquisition sensor 30.

[0081] In step S25, if the behavior detection unit 41f determines that any of the acquired measurement values ​​for each item exceeds the corresponding behavior threshold ("YES" in step S25), the factor detection unit 41k determines, based on the information acquired by the environmental information acquisition unit 41j, whether or not there was a factor in the surrounding environment of user Y that worsened user Y's emotions immediately before the measurement value that exceeded the behavior threshold was acquired (step S26).

[0082] The factor detection unit 41k recognizes the content of the voices uttered by people around the user Y from the information acquired by the environmental information acquisition unit 41j, for example by using a voice recognition system using an existing language model, and if the recognized content makes the user Y feel anger, it determines that there is a factor that will worsen the user Y's emotions.

[0083] When the factor detection unit 41k determines that there were no factors in the user Y's surrounding environment that would worsen the user Y's emotions immediately before obtaining a measurement value that exceeded the behavioral threshold ("NO" in step S26), it determines that user Y was unable to suppress the worsening of his / her emotions and performed an associated behavior related to the worsening of his / her emotions (step S27).

[0084] Also, in step S26, when the factor detection unit 41k determines that there is a factor in the surrounding environment of the user Y that worsens the emotion of the user Y ("YES" in step S26), it determines that the behavior performed by the user Y is not a related behavior caused by a decline in cognitive function, and proceeds to step S28. When it is determined that there is a factor in the surrounding environment of the user Y that worsens the emotion of the user Y, the factor detection unit 41k may determine that the related behavior was performed with a lower weighting than if there was no such factor. For example, when there is no factor that worsens the emotion of the user Y, the number of times of the relevant related behavior may be counted as 1, whereas when there is a factor that worsens the emotion of the user Y, the number of times may be counted as 0.5.

[0085] The processing of steps S25 to S27 is repeated until the biometric information of each item falls below the emotion threshold ("NO" in step S28). When the biometric information of each item falls below the emotion threshold ("YES" in step S28), the process returns to step S21, and the processing of steps S21 to S27 is repeated. In step S28, when the number of items whose measured values ​​exceed the emotion threshold falls below the number used in the determination in step S22, the process may return to step S21. In addition, in step S28, when a predetermined period has elapsed since the processing of step S23, that is, when a predetermined period has elapsed after the detection of the deterioration of the emotion of user Y, the process may return to step S21. Here, the predetermined period after the detection of the deterioration of the emotion is, for example, 10 minutes.

[0086] In the above-described embodiment, the emotion detection unit 41c acquires the information detected by the factor detection unit 41k, and in step S22, when the emotion detection unit 41c determines that the measurement values ​​of a predetermined number or more items exceed the emotion threshold and determines that there are no factors that may worsen the emotion of user Y, it may determine that user Y's emotion has worsened.

[0087] According to the above-described second embodiment, it is possible to generate cognitive function estimation information of user Y with high accuracy by narrowing down as much as possible to information and related behaviors when user Y's emotions deteriorate due to a decline in cognitive function.

[0088] In the above-mentioned first and second embodiments, the emotion threshold calculation unit 41b calculates the emotion threshold for each hour, but the unit of the time period for calculating the emotion threshold is not limited to this. For example, the emotion threshold calculation unit 41b may calculate the emotion threshold for each of three time periods, morning, afternoon, and night, or may calculate the emotion threshold for each time period during sleep and awake. This allows the emotion detection unit 41c to easily execute a process of determining whether the user Y has behaved differently from normal while appropriately identifying the difference in the user Y's behavior for each time period.

[0089] In the above-mentioned first and second embodiments, the first period T1 is one day, the second period T2 is one month, and the third period T3 is one year. However, this is not limited to this, and any periods may be set as long as the second period T2 is longer than the first period T1, and the third period T3 is longer than the second period T2.

[0090] In addition, by programming the functional configuration of the above-mentioned cognitive function estimation device 40A or 40B and incorporating it into a computer, it is also possible to construct a cognitive function estimation program that causes the computer to function as the cognitive function estimation device 40A or 40B.

[0091] The process executed by the cognitive function estimation device 40A or 40B described above can be expressed as a cognitive function estimation method executed by a cognitive function processing device. The cognitive function estimation method includes an emotion detection step of detecting a deterioration in the emotion of the user based on the biometric information of the user, a behavior detection step of detecting related behavior related to the deterioration in the emotion at a time point after the deterioration in the emotion of the user is detected, and an estimation step of estimating a decline in the cognitive function of the user based on the detection result in the emotion detection step and the detection result in the behavior detection step. [Explanation of symbols]

[0092] 1A, 1B Cognitive function estimation system 10 Biometric information measurement sensor 10a Pulse Meter 10b Blood Pressure Measuring Device 10c Skin potential measuring device 10d Autonomic nervous system measuring device 10e emotion recognition device 20 Behavioral information measurement sensor 20a Sound recording device 20b Acceleration sensor 30 Environmental information acquisition sensor 40A, 40B Cognitive function estimation device 41A, 41B CPU 41a Biometric information acquisition unit 41b Emotion threshold calculation unit 41c Emotion detection section 41d Behavior information acquisition unit 41e Behavioral threshold calculation unit 41f Behavior detection section 41g Short-term analysis department 41h Long-term analysis department 41i Display information generation section 41j Environmental Information Acquisition Department 41k Factor detection unit 42A, 42B storage section 42a Biometric information storage unit 42b Emotion Threshold Memory 42c Emotion detection information storage unit 42d Behavioral information storage unit 42e Behavioral threshold memory unit 42f Related behavior information storage section 42g Short-term analysis information storage unit 42h Long-term analysis information storage unit 42i Factor information storage section 43 Display section

Claims

1. an emotion detection unit that detects a deterioration in the emotion of the user based on biometric information of the user; A behavior detection unit that detects a related behavior related to the deterioration of the user's emotion at a time point after the deterioration of the user's emotion is detected; A cognitive function estimation device comprising: an estimation unit that estimates a decline in the user's cognitive function based on a detection result by the emotion detection unit and a detection result by the behavior detection unit.

2. The cognitive function estimation device according to claim 1 , wherein the estimation unit estimates a decline in the cognitive function of the user based on a time from when the emotion detection unit detects a deterioration in the user's emotion to when the behavior detection unit detects the related behavior.

3. The cognitive function estimation device according to claim 1 , wherein the estimation unit estimates the decline in the cognitive function of the user based on the number of times the related behavior is detected by the behavior detection unit within a predetermined period.

4. An environmental information acquisition unit that acquires information about the user's surrounding environment; a factor detection unit that detects a factor that is in the surrounding environment and that worsens the user's emotion based on the information acquired by the environmental information acquisition unit, The cognitive function estimation device according to claim 1 , wherein the estimation unit estimates a decline in the user's cognitive function based on a detection result by the emotion detection unit, a detection result by the behavior detection unit, and a detection result by the factor detection unit.

5. Cognitive function estimation device, An emotion detection function that detects a deterioration in the emotion of the user based on biometric information of the user; A behavior detection function that detects a related behavior related to the deterioration of the user's emotion at a time point after the deterioration of the user's emotion is detected; A cognitive function estimation program for realizing an estimation function that estimates a decline in the user's cognitive function based on the detection results by the emotion detection function and the detection results by the behavior detection function.

Images