JP2017012245A - Incontinence suppression method and incontinence suppression apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To further accurately suppress incontinence of a patient by taking account of a patient's condition as one aspect.SOLUTION: An incontinence suppression apparatus 40 extracts patterns of feature amounts representing common change tendencies when a patient has incontinence for each different condition of the patient classified by a specific feature amount out of a plurality of feature amounts on the basis of the plurality of feature amounts representing biological activities of the patient. Further, the incontinence suppression apparatus 40 estimates incontinence of the patient on the basis of the condition of the patient and the change tendencies of the feature amounts represented by the patterns according to the conditions, and, when estimating the incontinence of the patient, outputs an alert.SELECTED DRAWING: Figure 1

Description

  The disclosed technology relates to an incontinence suppressing method and an incontinence suppressing device.

  In recent years, a sensor is attached to the human body of a patient who has a tendency to incontinence, and the ultrasonic wave emitted from the sensor is reflected on the bladder and the bladder fullness is detected based on the reflection time until it returns to the sensor again. A sensing device that estimates urination time has been proposed.

  In this sensing device, a method of increasing the estimation accuracy of the urination time by adjusting the prediction parameter used when estimating the urination time from the difference between the actual urination time and the estimated urination time is used.

Special table 2013-523305 gazette

  However, even if the patient is the same, for example, if the patient's condition such as physical condition is different, the patient's actual urination time tends to vary. For this reason, when the urination time is estimated using a common prediction parameter without considering the patient's condition, the estimation accuracy of the urination time may be lowered.

  In one aspect, the disclosed technique aims to more accurately suppress incontinence of a patient by considering the patient's condition.

  In one aspect, the incontinence suppressing method causes the computer to incontinence the patient for each different state of the patient that can be distinguished by a specific feature amount out of the plurality of feature amounts based on the plurality of feature amounts indicating the life activity of the patient. The pattern of the feature amount showing the common change tendency is extracted. The incontinence suppression method causes the computer to predict the patient's incontinence based on the patient's condition and the change tendency of the feature amount indicated by the pattern corresponding to the condition, and warns when the patient's incontinence is predicted. Output.

  As one aspect, taking into account the patient's condition, there is an effect that the patient's incontinence can be more accurately suppressed.

It is a figure which shows an example of the change tendency of the feature-value for every measurement item according to sleep depth. It is a figure which shows an example of a structure of the incontinence suppression system which concerns on 1st Embodiment. It is a figure which shows an example of a feature-value list | wrist. It is a figure which shows an example of a structure in the case of implement | achieving the incontinence suppression apparatus which concerns on 1st Embodiment with a computer. It is a flowchart which shows an example of the flow of a feature-value list production | generation process. It is a flowchart which shows an example of the flow of a feature-value list initialization process. It is a flowchart which shows an example of the flow of an alert notification process. It is a figure which shows an example of a structure of the incontinence suppression system which concerns on 2nd Embodiment. It is a figure which shows an example of a structure in the case of implement | achieving the incontinence suppression apparatus which concerns on 2nd Embodiment with a computer. It is a flowchart which shows an example of the flow of a feature-value list update process.

  Hereinafter, an example of an embodiment of the disclosed technology will be described in detail with reference to the drawings. In addition, the same code | symbol may be provided to the component or process which bears the same function through all the drawings, and the overlapping description may be abbreviate | omitted suitably.

  The inventor attaches various sensors that measure the patient's biological information such as heart rate or respiratory rate to patients with incontinence, especially nocturnal urine tendency, and the biological information until the patient is incontinent for each measurement item. The changes in the data were acquired and the acquired biological information was analyzed.

  The measurement items related to the patient's life activity measured in the analysis are LF / HF, trunk movement, chest temperature, wrist temperature, temperature difference between chest and wrist, respiratory rate, wrist movement, heart rate, sleep These are posture, turnover frequency, and sleep depth.

  Here, LF / HF is a measurement item related to a patient's heart rate variability, and is measured by, for example, an ECG (Electrocardiograph) sensor used as an electrocardiograph. LF (Low Frequency) is an average of power spectra having a frequency exceeding 0.04 Hz and not more than 0.15 Hz among frequency power spectra obtained as a result of frequency conversion by performing frequency conversion on heart rate variability. The value. Further, HF (High Frequency) refers to an average value of power spectra having a frequency of more than 0.15 Hz and not more than 0.4 Hz in a power spectrum of frequencies with respect to heartbeat variability.

  LF is an index that reflects sympathetic nerve activity that works mainly when the body is active, and HF is an index that reflects parasympathetic nerve activity that works mainly when the body is in a relaxed state. It has been known. Therefore, LF / HF is used as an index indicating the degree of tension of the patient, and the value of LF / HF tends to decrease as the patient is in a relaxed state.

  The body movement of the trunk is a measurement item relating to the degree of body movement of the patient. The body movement of the trunk is represented by, for example, a value of an acceleration sensor that measures the movement of the patient's chest, and the value increases as the movement of the patient's body increases.

  The chest temperature is expressed, for example, as a value of the body surface temperature of the patient's chest, and the wrist temperature is expressed, for example, as the value of the body surface temperature of the patient's wrist.

  The temperature difference between the chest and the wrist is expressed as, for example, (chest temperature-wrist temperature), but may be (wrist temperature-chest temperature).

  The respiration rate is expressed as a value of a respiration sensor that detects the respiration rate per unit time (for example, 1 minute) of the patient from the movement of the chest of the patient, for example.

  The movement of the wrist is a measurement item related to the degree of body movement of the patient as well as the body movement of the trunk. The wrist movement is represented by, for example, a value of an acceleration sensor that measures the movement of the patient's wrist, and the value increases as the movement of the patient's hand increases.

  The heart rate is represented, for example, as a value of a heart rate sensor that detects a heart rate per unit time (for example, 1 minute) of the patient from the movement of the patient's pulse.

  The sleep posture is a measurement item related to the orientation of the patient's body during sleep, and is measured by, for example, a gyro sensor attached to the patient. Or you may measure from the direction of the gravitational acceleration acquired with an acceleration sensor.

  The turnover frequency is a measurement item calculated based on the sleep posture, and is represented, for example, by the number of turnovers performed by the patient per unit time (for example, 1 minute).

  The sleep depth is a measurement item related to the degree of sleep depth of the patient. It is known that the human sleep depth periodically repeats deepening and shallowing. A state in which the sleep depth is relatively shallow is called “REM sleep”, and the amount of activity of the body is decreasing, while the brain is active. Because the brain is already active during REM sleep, the patient may feel urinating and may want to go to the bathroom. When the patient's sleep depth is shallow, the patient can easily wake up, and the brain is already active, so the awakening after getting up is good.

  On the other hand, a state where the sleep depth is relatively deep is called “non-REM sleep”, and the amount of activity in both the body and the brain is lower than that in REM sleep. Non-REM sleep has lower brain activity compared to REM sleep, so even if you try to wake up a patient during non-REM sleep, the patient is difficult to wake up. There may be situations where feelings are recognized and awakening is not good.

  In addition, there is a tendency that a number of θ waves, which are brain waves having a frequency exceeding 4 Hz and about 8 Hz or less, appear during REM sleep compared to non-REM sleep. On the other hand, during non-REM sleep, there is a tendency for more δ waves, which are brain waves having a frequency of about 4 Hz or less, to appear compared to during REM sleep.

  Therefore, the patient's sleep depth can be detected by, for example, an electroencephalogram sensor that is mounted on the patient's head and detects the patient's brain waves, and the detected sleep depth is a threshold value indicating the boundary between REM sleep and non-REM sleep. Based on the comparison result, the sleep depth is classified into two stages, “deep” and “shallow”. Or you may obtain | require sleep depth from the frequency of a motion using the acceleration of a wrist.

  As a biometric information analysis procedure, first, for example, biometric information during sleep of a patient within a predetermined period such as one month is measured for each measurement item, and if the patient is incontinent during sleep, the incontinence time of the patient is determined. Record. Then, the average incontinence time of the patient within a predetermined period from the recorded incontinence time of the patient is obtained in advance.

  Then, for each measurement item other than the sleep depth, from the average value of the measurement items measured for 60 minutes before the incontinence of the patient on the day of incontinence, for 60 minutes before the average incontinence time of the patient on the day without incontinence. The average value of the measured values of the measurement items is subtracted, and a difference is calculated for each measurement item. Hereinafter, this difference is referred to as a feature amount, and the feature amount for each measurement item is calculated for each day of incontinence.

  Then, each feature amount of each measurement item on the day of incontinence is classified into two patterns, when the sleep depth at the time of incontinence is deep (represented as pattern 1) and when it is shallow (represented as pattern 2). And in each pattern, the measurement item which has the feature-value which shows the common change tendency among each measurement item is extracted.

  Specifically, if any feature value of a certain measurement item is a positive value in any pattern, the measurement item is extracted from the pattern as a measurement item indicating a change tendency of the feature amount. . On the other hand, if any feature value of a certain measurement item is a negative value in any pattern, the measurement item is extracted from the pattern as a measurement item indicating a change tendency of the feature amount.

  As an example, since each pattern is classified according to the sleep depth, the sleep depths of the patients included in the same pattern all indicate the same sleep depth.

  FIG. 1 is a diagram illustrating an example of the change tendency of the feature amount of each measurement item according to the sleep depth of four patients represented by patients 1 to 4 analyzed as described above.

  Here, “↑” indicates that there is a common change tendency that the feature value of the corresponding measurement item increases in each pattern, and “↓” indicates the feature value of the corresponding measurement item in each pattern. It shows that there is a common trend of change.

  Also, the blank in FIG. 1 indicates that a common change tendency could not be found because there are both feature quantities that rise or fall for the corresponding measurement item.

  Therefore, the feature amount in pattern 1 of patient 1 is LF / HF, respiratory rate, heart rate, and sleep depth, and the feature amount in pattern 2 of patient 1 is LF / HF, temperature difference between chest and wrist, respiratory rate. , Heart rate, and sleep depth.

  The feature amount in pattern 1 of patient 2 is LF / HF, chest temperature, wrist temperature, respiratory rate, heart rate, and sleep depth. The feature amount in pattern 2 of patient 2 is LF / HF, chest Temperature, respiratory rate, heart rate, sleep posture, and sleep depth. Hereinafter, the feature amounts are also shown for the patient 3 and the patient 4.

  The inventors were able to obtain some knowledge from the analysis results shown in FIG.

  The first finding is that there are feature quantities that show the same change tendency between different patterns and feature quantities that show different change trends.

  For example, in the analysis result of patient 1, LF / HF and heart rate show the same change tendency between pattern 1 and pattern 2, but the respiratory rate and sleep depth are between pattern 1 and pattern 2. Shows different change trends.

  The second finding is that the characteristic amount before incontinence varies from patient to patient.

  For example, in the case of the pattern 2 of the patient 1, there is a change tendency in which the temperature difference between the chest and the wrist of the patient 1 increases. In the case of the pattern 2 of the patient 2, the temperature difference between the chest and the wrist of the patient 2 is observed. There is no particular change trend.

  The third finding is that even if the pattern is the same, the change tendency of the feature amount is different for each patient.

  For example, in the case of the pattern 2 of the patient 1, a change tendency in which the respiration rate of the patient 1 decreases is seen, but in the case of the pattern 2 of the patient 2, a change tendency in which the respiration rate of the patient 2 increases is seen. .

  From the above, before the incontinence of the patient, the inventors have a tendency of change of the feature amount that appears in common with respect to the specific measurement item according to the sleep depth, and the change tendency of the specific measurement item and the feature amount is We found that it was different for each patient.

  For example, patient 1 tends to increase LF / HF and breathing rate and decrease heart rate when incontinence is performed at a deep sleep depth, and LF / HF and chest when incontinence occurs at a shallow sleep depth. There is a tendency for the temperature difference between the wrist and wrist to increase, and the respiratory rate and heart rate to decrease.

  Therefore, if a measurement item corresponding to a feature quantity that shows a common change tendency in any pattern among the measurement items shows the same change tendency as the change tendency indicated by the feature quantity, the patient may be incontinent. Something can be detected in advance before the patient is incontinent.

  Based on the knowledge shown above, hereinafter, an incontinence suppressing device that predicts patient incontinence based on a change tendency of biological information according to the patient's sleep depth and suppresses incontinence of the patient will be described.

(First embodiment)
FIG. 2 is a diagram illustrating an example of the configuration of the incontinence suppression system 1 according to the first embodiment.

  The incontinence suppression system 1 includes a wearable sensor 10, an information terminal 30, and an incontinence suppression device 40. The wearable sensor 10 and the information terminal 30, the information terminal 30 and the incontinence suppressing device 40, and the incontinence suppressing device 40 and the wearable sensor 10 are connected by a communication line.

  Although FIG. 2 shows an example in which one information terminal 30 is connected to the incontinence suppressing device 40, a plurality of information terminals 30 can be connected to the incontinence suppressing device 40. Further, the number of wearable sensors 10 that can be connected to the information terminal 30 and the incontinence suppressing device 40 is not limited.

  In the example of FIG. 2, the wearable sensor 10 is connected to the incontinence suppressing device 40 via the information terminal 30, but the wearable sensor 10 may be connected to the incontinence suppressing device 40.

  There is no restriction on the line type of the communication line, and it may be any line type such as an Internet line, an intranet line, or a dedicated line, and the connection form is also wired, wireless, or a mixture of wired and wireless. Any connection form may be used.

  The wearable sensor 10 is an electronic device that measures a patient's biological information and outputs a warning based on an instruction from the incontinence suppressing device 40, and is worn or carried by the patient.

  The wearable sensor 10 includes a transmission unit 11, a urine sensor 12, an acceleration sensor 13, a temperature sensor 14, a respiration sensor 15, a heart rate sensor, a gyro sensor 17, an electroencephalogram sensor 18, and an ECG sensor 19. The wearable sensor 10 includes an alert receiver 20 and a speaker 21.

  The transmitting unit 11 of the wearable sensor 10 includes a sensor data receiving unit 31 of the information terminal 30, a urine sensor 12, an acceleration sensor 13, a temperature sensor 14, a respiratory sensor 15, a heart rate sensor, a gyro sensor 17, an electroencephalogram sensor 18, and an ECG sensor. 19 is connected.

  The urine sensor 12 detects incontinence of the patient, and outputs incontinence detection data notifying the detection of incontinence to the transmission unit 11.

  The acceleration sensor 13 measures the acceleration of the movement of the chest of the patient and the acceleration of the movement of the wrist of the patient, and outputs the acceleration data for each part to the transmission unit 11. The acceleration data indicating the movement of the patient's chest represents the biological information corresponding to the body movement of the trunk among the measurement items. The acceleration data indicating the movement of the patient's wrist represents biological information corresponding to the movement of the wrist among the measurement items.

  The temperature sensor 14 measures the temperature of the patient's chest and the temperature of the patient's wrist, calculates the temperature difference between the chest and the wrist from the measurement result, and outputs the temperature data and temperature difference data for each part to the transmission unit 11. To do.

  The respiration sensor 15 detects, for example, the movement of the chest of the patient, measures the respiration rate of the patient per unit time, and outputs respiration data indicating the respiration rate of the patient to the transmission unit 11. In addition, respiration data represents the biometric information corresponding to the respiration rate among measurement items.

  For example, the heart rate sensor 16 detects the movement of the patient's pulse, measures the heart rate of the patient per unit time, and outputs the heart rate data indicating the heart rate of the patient to the transmission unit 11. The heartbeat data represents biological information corresponding to the heart rate among the measurement items.

  The gyro sensor 17 detects the orientation of the patient's body, measures which direction the patient is facing, and calculates the number of movements of the patient's body per unit time using the measurement result. The gyro sensor 17 outputs gyro data indicating the orientation of the patient's body and roll-over data indicating the number of movements of the patient's body per unit time to the transmission unit 11.

  The gyro sensor 17 indicates a value of “0”, for example, when the patient is in the supine state, and is a value that increases in the “+” direction as the patient's body rotates leftward from the supine state as viewed from the patient. Conversely, a value that decreases in the “−” direction as it rotates in the right direction. This makes it possible to estimate the orientation of the patient's body from the gyro data.

  Note that the gyro data represents the biological information corresponding to the sleep posture among the measurement items, and the turnover data represents the biological information corresponding to the turnover frequency among the measurement items.

  The electroencephalogram sensor 18 detects a potential difference generated in the patient's head as a result of brain activity, and measures the patient's electroencephalogram. At this time, the electroencephalogram sensor 18 sets, for example, “+” for a situation where more θ waves appear than δ waves, and “−” for situations where more δ waves appear compared to θ waves. The electroencephalogram data in which the length is digitized is output to the transmitter 11. The electroencephalogram data represents biological information corresponding to the sleep depth among the measurement items.

  The ECG sensor 19 detects the patient's heart rate variability, measures the activity state of the patient's autonomic nerve, that is, LF / HF, and outputs ECG data indicating the patient's LF / HF to the transmitter 11. The ECG data represents biological information corresponding to LF / HF among the measurement items. The heart rate may be obtained from the RR interval (R wave interval) of the ECG sensor.

  Each said sensor 12-19 measures a patient's biometric information, for example for every predetermined space | interval, and outputs the various data (biometric data) corresponding to each sensor 12-19 to the transmission part 11. FIG. There is no particular limitation on the measurement interval of the biological information in each of the sensors 12 to 19, but the amount of data per unit time increases as the measurement interval of the biological information becomes shorter. Can be accurately predicted.

  The sensors 12 to 19 do not need to be built in the wearable sensor 10, and the sensors 12 to 19 may be separated from the wearable sensor 10.

  The transmission unit 11 receives the biometric data output from each of the sensors 12 to 19, adds a date and time when the biometric data is received, and a patient ID (identification) for identifying the patient for each biometric data, It transmits to the information terminal 30. The date and time when the biometric information is measured may be added to the biometric data by the sensors 12 to 19, and the patient ID is added to the biometric data at the information terminal 30 that is the transmission destination of the biometric data. You may make it do.

  The alert receiving unit 20 of the wearable sensor 10 is connected to the alert output unit 44 of the incontinence suppressing device 40 and the speaker 21.

  Upon receiving an alert output instruction from the alert output unit 44, the alert receiving unit 20 outputs audio data corresponding to the warning sound to the speaker 21.

  When the speaker 21 receives the voice data from the alert receiver 20, the speaker 21 outputs a voice corresponding to the voice data, that is, a warning sound. Note that the speaker 21 does not have to be built in the wearable sensor 10, and the speaker 21 may be separated from the wearable sensor 10.

  On the other hand, the information terminal 30 is installed in a patient's house, for example, and executes processing related to input / output of various information related to patient incontinence and output of warnings.

  The information terminal 30 includes a sensor data receiving unit 31, a sensor data transmitting unit 32, an input unit 33, an information terminal transmitting unit 34, a speaker 35, an alert receiving unit 36, and a lighting control unit 37.

  The sensor data reception unit 31 is connected to the transmission unit 11 and the sensor data transmission unit 32 of the wearable sensor 10, and outputs the patient biometric data received from the transmission unit 11 to the sensor data transmission unit 32.

  The sensor data transmission unit 32 is connected to the sensor data reception unit 31 and the collection recording unit 41 of the incontinence suppressing device 40, and transmits the biological data received from the sensor data reception unit 31 to the collection recording unit 41.

  The input unit 33 is connected to the information terminal transmission unit 34, receives input information input from an input device such as a touch panel, a keyboard, and a mouse, and outputs the received input information to the information terminal transmission unit 34.

  The information terminal transmission unit 34 is connected to the input unit 33 and the collection recording unit 41 of the incontinence suppressing device 40, and transmits the input information received from the input unit 33 to the collection recording unit 41.

  The alert receiving unit 36 is connected to the alert output unit 44 of the incontinence suppressing device 40, and the speaker 35 and the illumination control unit 37 of the information terminal 30. When the alert receiving unit 36 receives an alert output instruction from the alert output unit 44, the alert receiving unit 36 outputs audio data corresponding to the warning sound to the speaker 35 and outputs an illumination control instruction to the illumination control unit 37.

  When the speaker 35 receives the sound data from the alert receiver 36, the speaker 35 outputs a sound corresponding to the sound data, that is, a warning sound.

  When the illumination control instruction is received from the alert receiving unit 36, the illumination control unit 37 turns on the light of a lighting device (not shown) installed in the room where the patient and the guardian of the patient are sleeping, for example.

  On the other hand, the incontinence suppressing device 40 includes a collection recording unit 41, an extraction unit 42, a determination unit 43, an alert output unit 44, and a storage unit 45 that stores a sensor data DB (Database) 48 and a feature amount list DB 49.

  The collection recording unit 41 is connected to the sensor data transmission unit 32 and the information terminal transmission unit 34 of the information terminal 30 and accepts patient biometric data and input information. Furthermore, the collection recording unit 41 is connected to the extraction unit 42, the determination unit 43, and the storage unit 45, and records the received biological data in the sensor data DB 48 in time series. At this time, the collection recording unit 41 refers to the patient ID added to the biometric data, associates the biometric data with the patient ID, and records them in the sensor data DB 48. Further, the collection / recording unit 41 outputs the received input information to the extraction unit 42 or the determination unit 43, and provides information necessary for the processing of the extraction unit 42 or the determination unit 43.

  The extraction unit 42 is connected to the collection recording unit 41 and the storage unit 45 and acquires, for example, the patient biometric data specified by the information terminal 30 from the sensor data DB 48. And the extraction part 42 classify | categorizes the calculated feature-value into each pattern with the case where a sleep depth is deep, and a shallow case, while calculating the feature-value for every measurement item from the acquired patient biometric data, and each pattern From these, measurement items indicating a change tendency with common feature values are extracted.

  At this time, the extraction unit 42 responds to the extracted measurement item according to the change tendency of the feature quantity indicating whether the feature quantity tends to increase or the tendency to decrease, and the degree of change tendency of the feature quantity. Set the threshold value. And the extraction part 42 matches a threshold value with the extracted measurement item for every patient pattern, and memorize | stores it in feature-value list DB49. The association between the measurement item and the threshold value in each pattern of the patient is referred to as a feature amount list, and the feature amount list is generated for each patient and stored in the feature amount list DB 49.

  FIG. 3 is a diagram illustrating an example of the feature amount list of the patient 1 illustrated in FIG. As shown in FIG. 3, the feature quantity list S classifies, for example, measurement items and threshold values for each pattern. For example, in Pattern 1, since LF / HF is represented by a positive value of “3.0”, LF / HF tends to increase and the absolute value of the threshold is “3.0”. It is shown that. In Pattern 1, since the heart rate is represented by a negative value of “−1.8”, the heart rate tends to change, and the absolute value of the threshold is “1.8”. Is shown.

  The determination unit 43 is connected to the collection recording unit 41, the alert output unit 44, and the storage unit 45. The determination unit 43 acquires, for example, the biological data of the sleeping patient designated by the information terminal 30 from the sensor data DB 48, and also acquires the feature amount list S corresponding to the sleep depth of the patient from the feature amount list DB 49. To do.

  And the determination part 43 calculates a feature-value for every measurement item using the biological data of the patient in sleep. The determination unit 43 matches the change amount of the threshold value of the measurement item included in the feature amount list S with the feature amount of each measurement item corresponding to the measurement item included in the feature amount list S among the calculated feature amounts. If the threshold is exceeded, an alert output instruction is output to the alert output unit 44.

  The alert output unit 44 is connected to the determination unit 43, the alert reception unit 36 of the information terminal 30, and the alert reception unit 20 of the wearable sensor 10, and sends the alert output instruction received from the determination unit 43 to the alert reception units 20 and 36. Output.

  Next, FIG. 4 shows an example of a configuration diagram when the incontinence suppressing device 40 is realized by a computer.

  The computer 100 includes a CPU 102, a memory 104, and a nonvolatile storage device 106. The CPU 102, the memory 104, and the nonvolatile storage device 106 are connected to each other via a bus 108. The computer 100 also includes an input / output (I / O) 110 for connecting the input device 112 and the communication device 114 to the computer 100 and transmitting / receiving data to / from each other. The I / O 110 is connected to the bus 108. Is done.

  The input device 112 includes an input device for giving an instruction to the computer 100 by an operator of the computer 100 such as a keyboard and a mouse. The input device 112 includes a reading device for reading data recorded on the recording medium 116 such as a CD-ROM or a flash memory. Note that the input device 112 is not necessarily connected to the I / O 110.

  The communication device 114 includes an interface for connecting to a communication line, and performs data transmission / reception with the wearable sensor 10 and the information terminal 30 connected to the communication line, for example.

  The storage device 106 can be realized by an HDD (Hard Disk Drive) or a flash memory.

  The storage device 106 stores an incontinence suppression program 120 for causing the computer 100 to function as the incontinence suppression device 40 shown in FIG. The incontinence suppression program 120 stored in the storage device 106 includes a collection recording process 122, an extraction process 124, a determination process 126, and an alert output process 128.

  The CPU 102 reads out the incontinence suppression program 120 from the storage device 106 and expands it in the memory 104, and executes each process included in the incontinence suppression program 120.

  The CPU 102 reads out the incontinence suppression program 120 from the storage device 106, expands it in the memory 104, and executes the incontinence suppression program 120, whereby the computer 100 operates as the incontinence suppression device 40 shown in FIG.

  Further, when the CPU 102 executes the collection recording process 122, the computer 100 operates as the collection recording unit 41 illustrated in FIG. Further, when the CPU 102 executes the extraction process 124, the computer 100 operates as the extraction unit 42 illustrated in FIG. Further, when the CPU 102 executes the determination process 126, the computer 100 operates as the determination unit 43 illustrated in FIG. Further, when the CPU 102 executes the alert output process 128, the computer 100 operates as the alert output unit 44 shown in FIG.

  Furthermore, the sensor data DB 48 is stored in the memory 104 by the CPU 102 developing the sensor data information stored in the sensor data information storage area 140 in the memory 104. Further, the CPU 102 expands the feature quantity list information stored in the feature quantity list information storage area 142 in the memory 104, whereby the feature quantity list DB 49 is stored in the memory 104.

  The computer 100 can be realized by, for example, a semiconductor integrated circuit, more specifically, an ASIC (Application Specific Integrated Circuit) or the like.

  Next, the operation of the incontinence suppressing device 40 according to the first embodiment will be described. For example, the incontinence suppressing device 40 according to the first embodiment executes a feature list generation process based on an instruction from the information terminal 30.

  FIG. 5 is a flowchart showing an example of the flow of feature quantity list generation processing of the incontinence suppressing device 40. The sensor data DB 48 of the incontinence suppressing device 40 is assumed to store biological data for a predetermined period for each patient in advance.

  In step S10, the extraction unit 42 calculates feature amounts from, for example, biological data of any patient within a predetermined period, classifies the calculated feature amounts into two patterns according to the patient's sleep depth, and stores the feature amount list. The feature amount list initialization process for initializing is executed.

  FIG. 6 is a flowchart illustrating an example of the flow of feature quantity list initialization processing in step S10.

  In step S200, the extraction unit 42 refers to the incontinence detection data stored in the sensor data DB 48, and is recorded in a period from 60 minutes before the patient's incontinence to immediately before the incontinence time for each incontinence of the patient within a predetermined period. The obtained biological data is acquired from the sensor data DB 48. Hereinafter, the period from 60 minutes before the patient's incontinence to immediately before the incontinence time is referred to as the counting period.

Then, the extraction unit 42 calculates an average value for each measurement item of the biometric data during the total period for each total period, and calculates incontinence biometric data A _i [d]. That is, the incontinence biometric data A _i [d] indicates the biometric data of the patient before the patient is incontinent.

  Here, the variable i (= 1, 2,...) Is a number indicating the measurement item, and the variable d (= 1, 2,...) Is a number for identifying the counting period. For example, LF / HF is associated with i = 1, and body movement of the trunk is associated with i = 2. Further, when d = 1, the oldest aggregation period within the predetermined period, and when d = 2, the aggregation period is associated with the aggregation period next to the aggregation period indicated by d = 1 within the predetermined period. Is done.

  There is no limitation on the association between the measurement item and the variable i and the association between the total period and the variable d, and any association may be performed. For example, i = 2 may be associated with LF / HF, and d = 1 may be associated with the latest aggregation period within a predetermined period. The predetermined period here is assumed to be a period of about one month, for example, but there is no particular limitation on the period. Hereinafter, the variable i is referred to as a measurement item i, and the variable d is referred to as a total period d.

  Furthermore, the extraction unit 42 calculates the average incontinence time of the patient in a predetermined period.

  In step S <b> 210, the extraction unit 42 refers to the incontinence detection data stored in the sensor data DB 48, and acquires from the sensor data DB 48 patient biometric data for each day that the patient within the predetermined period is not incontinent. Specifically, the extraction unit 42 records the biometric data recorded in a period from 60 minutes before the average incontinence time calculated in step S200 to immediately before the average incontinence time every day when the patient within the predetermined period is not incontinent. Is acquired from the sensor data DB 48.

Then, the extraction unit 42 calculates reference biological data B _i that is obtained by averaging all the biological data within the predetermined period acquired in step S210 for each measurement item i. That is, the reference biometric data B _i indicates the biometric data of the patient when the patient does not incontinence.

In step S200 and step S210, the biometric data acquisition period is set to 60 minutes as an example. However, this value is an example, and the biometric data acquisition period may be longer or shorter than 60 minutes. . However, the biometric data acquisition period in step S200 and step S210 is preferably the same. This is because if the biometric data acquisition period is different, a significant difference between the incontinence biometric data A _i [d] and the reference biometric data B _i may not be correctly evaluated.

In step S220, the extraction unit 42 uses the incontinence biometric data A _i [d] calculated in step S200 and the reference biometric data B _i calculated in step S210, and features for each measurement item i in each counting period. C _i [d] is calculated by equation (1).

(Equation 1)
C _i [d] = A _i [d] −B _i (1)

That is, the characteristic amount C _i [d] is biological information obtained by removing a patient's inherent biological tendency such as a low respiratory rate even when there is no incontinence from the biological tendency of the patient. Represents information changes.

The extraction unit 42 stores the calculated feature amount C _i [d] in, for example, a predetermined area of the memory 104.

Note that the feature quantity C _i [d] for each measurement item i may be calculated by the expression (2) instead of the expression (1), and the patient's inherent characteristic from the incontinence biometric data A _i [d]. The patient's feature amount C _i [d] may be calculated using another arithmetic expression that removes the biological tendency.

(Equation 2)
C _i [d] = B _i −A _i [d] (2)

In step S230, the extracting unit 42 refers to the depth of sleep incontinence immediately before the biometric data used for calculation of incontinence Biometric data A _i [d], among incontinence Biometric data A _i [d], the patient Incontinence biometric data A _i [d ₁ ] when the sleep depth is deep is detected. Here, aggregation period d ₁ is one or more of any integer.

Then, the extraction unit 42 uses the feature amount C _i [d ₁ ] corresponding to the incontinence biometric data A _i [d ₁ ] belonging to the pattern 1 where the patient's sleep depth is deep, and the feature amount list S1 in the pattern 1 is obtained. initialize. Specifically, the extraction unit 42 generates a feature amount list S1 in which the measurement item i and the feature amount C _i [d ₁ ] are associated with each other, and stores the feature amount list S1 in a predetermined area of the memory 104, for example. Note that the feature quantity list S1 generated in step S230 is a provisional feature quantity list S1.

In step S240, the extraction unit 42 refers to the depth of sleep incontinence immediately before the biometric data used for calculation of incontinence Biometric data A _i [d], among incontinence Biometric data A _i [d], the patient Incontinence biometric data A _i [d ₂ ] when the sleep depth is shallow is detected. Here, the total period d ₂ is one or more of any integer.

Then, the extraction unit 42 uses the feature quantity C _i [d ₂ ] corresponding to the incontinence biometric data A _i [d ₂ ] belonging to the pattern 2 where the patient's sleep depth is shallow, and extracts the feature quantity list S2 in the pattern 2. initialize. Specifically, the extraction unit 42 generates a feature quantity list S2 in which the measurement item i and the feature quantity C _i [d ₂ ] are associated with each other, and stores the feature quantity list S2 in a predetermined area of the memory 104, for example. Note that the feature quantity list S2 generated in step S240 is a temporary feature quantity list S2.

  Thus, the feature amount list initialization process in step S10 in FIG. 5 ends, and the initialized feature amount list S1 and feature amount list S2 are stored in a predetermined area of the memory 104.

Next, processing after step S20 in FIG. 5 will be described. In step S20 and subsequent steps, the measurement item i indicating a different change tendency of the feature value C _i [d] is removed from the feature value list S1 generated in step S230 and the feature value list S2 generated in step S240. An amount list S1 and a feature amount list S2 are generated.

  First, in step S <b> 20, the extraction unit 42 initializes the counting period d to “0” and stores it in, for example, a predetermined area of the memory 104.

  In step S30, the extraction unit 42 acquires the aggregation period d from the memory 104, and increases (increments) the aggregation period d by one. Then, the extraction unit 42 stores the updated variable d in, for example, a predetermined area of the memory 104.

In step S <b> 40, the extraction unit 42 acquires a feature amount C _i [d] corresponding to the current value of the total period d from the memory 104. Note that the feature amount C _i [d] is calculated in step S220 in FIG.

In step S50, the extraction unit 42 refers to the sleep depth immediately before incontinence of the biometric data used to calculate the incontinence biometric data A _i [d] corresponding to the acquired feature amount C _i [d], and calculates the total period Obtain the patient's depth of sleep at d. And the extraction part 42 determines whether the patient's sleep depth in the total period d is deep, transfers to step S60 in the case of an affirmative determination, and transfers to step S70 in the case of a negative determination.

Note that the extraction unit 42 stores the acquired feature amount C _i [d] and the patient's sleep depth in the counting period d in association with each other, for example, in a predetermined area of the memory 104. That is, the extraction unit 42 records whether the acquired feature value C _i [d] belongs to the pattern 1 or the pattern 2.

In step S <b> 60, the extraction unit 42 acquires a feature amount list S <b> 1 that is a feature amount list when the sleep depth is deep from the memory 104. Whether extraction unit 42, the feature amount C _i [d _1] contained in the feature list S1, the feature amount C _i [d] obtained in step S40, the measurement item i show different change tendency in the presence Determine whether. Then, if there is a corresponding measurement item i, the extraction unit 42 removes the measurement item i from the feature amount list S1.

Specifically, the extraction unit 42 compares the feature quantity C _i [d ₁ ] included in the feature quantity list S1 with the sign of the feature quantity C _i [d] acquired in step S40 for each measurement item i, The measurement item i having a different sign of the feature amount and the feature amount C _i [d ₁ ] of the measurement item i are removed from the feature amount list S1.

  For example, the feature amount related to the chest temperature included in the feature amount list S1 is “1.2”, and the feature amount related to the chest temperature acquired in step S40 is “−0.8”. In this case, since the sign of one feature quantity is positive and the sign of the other feature quantity is negative, the measurement item related to the chest temperature and the feature quantity “1.2” are removed from the feature quantity list S1.

  The extraction unit 42 stores the feature quantity list S1 updated in this way as a new feature quantity list S1, for example, in a predetermined area of the memory 104.

In step S <b> 70, the extraction unit 42 acquires a feature amount list S <b> 2 that is a feature amount list when the sleep depth is shallow from the memory 104. Extraction unit 42, similarly to the step S60, the feature amount and C _i [d _2] contained in the feature list S2, measurement items indicating characteristic quantity and C _i [d] acquired in step S40, a different change tendency in It is determined whether i exists. Then, if there is a corresponding measurement item i, the extraction unit 42 removes the measurement item i from the feature amount list S2.

  The extraction unit 42 stores the updated feature quantity list S2 as a new feature quantity list S2, for example, in a predetermined area of the memory 104.

In step S80, the extraction unit 42 determines whether or not there is a feature amount C _i [d] corresponding to the total period d that has not yet been acquired in step S40. If the determination is affirmative, the extraction unit 42 proceeds to step S30. Transition. Then, the aggregation period d is updated in step S30, and the processes in steps S30 to S80 are repeated, so that the feature quantity list S1 and the feature quantity list S1 in which only the measurement item i having a common change tendency of the feature quantity C _i [d] remains are obtained. A quantity list S2 is generated.

On the other hand, if the determination process in step S80 is negative, that is, if all the feature values C _i [d] in the predetermined period are acquired from the memory 104 in step S40, the process proceeds to step S90.

In step S90, the extraction unit 42 sets a threshold value for the feature value C _i [d ₁ ] of each measurement item i included in the feature value list S1. Specifically, the extraction unit 42, for each measurement item i included in the feature list S1, of all the feature quantity C _i [d] included in the pattern 1, the feature quantity absolute value of C _i [d] The feature quantity C _i [d] having the smallest is extracted. Then, the extraction unit 42 sets the extracted feature quantity C _i [d] in the feature quantity list S1 as a threshold value of the feature quantity related to the measurement item i instead of the feature quantity C _i [d ₁ ].

Furthermore, in step S100, the extraction unit 42 sets a threshold value for the feature value C _i [d ₂ ] of each measurement item i included in the feature value list S2. Specifically, the extraction unit 42, for each measurement item i included in the feature list S2, among all the feature amount C _i [d] included in the pattern 2, wherein the absolute amount of C _i [d] The feature quantity C _i [d] having the smallest is extracted. Then, the extraction unit 42 sets the extracted feature value C _i [d] in the feature value list S2 as a feature value threshold for the measurement item i instead of the feature value C _i [d ₂ ].

Note that whether the feature amount C _i [d] is included in the pattern 1 or the pattern 2 is determined based on the feature amount C _i [d], the sleep depth of the patient stored in the memory 104 in step S50. Refer to the association.

Further, as an example, among the patterns 1 or all of the feature amount C _i contained in the pattern 2 [d], the absolute value of the minimum of the feature amount C _i of the feature C _i [d] in the measurement item i [d ] Is set as the threshold value of the feature amount related to the measurement item i, but the threshold value setting method is not limited to this. For example, an average value of all the feature values C _i [d] included in the pattern 1 or the pattern 2 for each measurement item i may be set as a feature value threshold value for the measurement item i. Further, of the pattern 1 or all of the feature amount C _i contained in the pattern 2 [d], the maximum feature quantity C _i [d] is the absolute value of the feature amount C _i [d] in the measurement item i, measurement items You may set to the threshold-value of the feature-value regarding i.

  In step S110, the extraction unit 42 stores the feature amount list S1 generated in step S90 and the feature amount list S2 generated in step S100 in the feature amount list DB 49.

  The feature quantity list generation process shown in FIG. By executing this feature quantity list generation process for each patient, the feature quantity list S1 and the feature quantity list S2 for each patient are stored in the feature quantity list DB 49.

  Note that after generating the feature quantity lists S1 and S2, a sensor that detects biometric data related to the measurement item i that is not included in the feature quantity lists S1 and S2 may be removed from the wearable sensor 10.

  Next, based on the created feature quantity list S1 and feature quantity list S2, it is determined whether or not there is a sign of incontinence in a sleeping patient, and an alert notification process for outputting a warning when incontinence of the patient is predicted Will be described. Here, although the example which performs alert notification processing with respect to the sleeping patient instruct | indicated with the information terminal 30 is demonstrated, it is not restricted to this, For example, with respect to all the sleeping patients which are collecting biometric data Alert notification processing may be executed.

  FIG. 7 is a flowchart showing an example of the alert notification processing flow of the incontinence suppressing device 40. The sensor data DB 48 of the incontinence suppressing device 40 stores biological data within a predetermined period for each patient in advance, and the feature amount list DB 49 includes a feature amount list S1 and a feature amount list corresponding to the sleep depth for each patient. Assume that S2 is stored in advance. When there is no need to distinguish between the feature quantity list S1 and the feature quantity list S2, the feature quantity list S1 is referred to as a feature quantity list S.

  The CPU 102 of the incontinence suppressing device 40 includes a calendar function for managing time, for example. Each process included in the incontinence suppression program 120 can acquire time information from the CPU 102 by using, for example, a predetermined API (Application Programming Interface).

  First, in step S <b> 300, the determination unit 43 acquires the patient ID of the patient instructed by the information terminal 30 from the collection recording unit 41. And the determination part 43 acquires the biometric data matched with designated patient ID from sensor data DB48.

  At this time, the determination unit 43 acquires, from the sensor data DB 48, biometric data recorded in a period from the time (current time) when step S300 is executed to 60 minutes before the time. The current time may be acquired from the CPU 102 using, for example, a predetermined API. Note that the current time acquisition method is not limited to this. For example, the current time may be acquired from a time server (not shown) connected to the communication line.

Determination unit 43 calculates the average value of the acquired biometric data for each measurement item i, and calculates the current biometric data A _i. In other words, the current biological data A _i indicates the current biological data of the sleeping patient.

Then, the determination unit 43 (1) by replacing _{A i} to [d] in the current biometric data _{A i} of equation from the obtained biometric data at step S300, the calculating the feature amount _{C i} of each measurement item i. The reference biometric data B _i to be used for calculating the feature quantity C _i may be used reference biometric data B _i calculated in step S210 in FIG. 6.

The feature amount C _i (current feature amount C _i ) calculated in this way can be regarded as indicating a change tendency of the current life activity of the sleeping patient.

Here, the biometric data acquisition period is set to 60 minutes, but the value of 60 minutes is an example, and the biometric data acquisition period may be longer or shorter than 60 minutes. However, it is preferable to make it the same as the biometric data acquisition period used in step S200 and step S210. This is because, when the feature amount of threshold for the measurement item i included in the feature lists S1, S2, the acquisition period of the biometric data used to calculate the current feature amount C _i of the patient, different, in the subsequent processing , there may be a case that the threshold and not currently accurately compare the feature amount C _i.

In step S310, the judgment unit 43, the current of the biometric data used for calculation of the feature amount C _i, with reference to the EEG data in the most recent time, it obtains the current sleep depth of the patient. And the determination part 43 determines whether a patient's present sleep depth is deep, and transfers to step S320 in negative determination.

  Note that the sleep depth of the patient can also be determined from biological data other than the electroencephalogram data. For example, since the movement of the wrist decreases as the sleep depth of the patient increases, the sleep depth of the patient can be determined from acceleration data indicating the movement of the wrist. Similarly, since the LF / HF value decreases as the patient's sleep depth increases, it is also possible to determine the patient's sleep depth from ECG data indicating LF / HF. In this case, the electroencephalogram sensor 18 can be omitted from the wearable sensor 10.

In step S320, which is a transition destination when it is determined that the sleep depth of the patient is shallow, the determination unit 43 acquires the feature amount list S2 of the patient from the feature amount list DB 49. Then, the determination unit 43, among the current feature quantity C _i of patients calculated in step S300, the current and the feature amount C _i about measurement items i included in the feature list S2, which has been acquired, is included in the feature list S2 Measurement The threshold value of the feature amount related to the item i is compared for each measurement item i. Then, the determination unit 43, for all the measurement item i included in the feature list S2, the change trend of the current feature quantity C _i matches a feature amount of change trends in the measurement item i included in the feature list S2 In addition, it is determined whether or not the current feature amount C _i exceeds a threshold value. Note that “the current feature value C _i exceeds the threshold value” means that the absolute value of the current feature value C _i exceeds the absolute value of the threshold value. If the determination in step S320 is affirmative, the process proceeds to step S340.

  The feature amount list S2 indicates a change tendency of the feature amount that commonly appears in the patient's body as a sign when the patient is incontinent when the sleep depth of the patient is shallow. Therefore, the determination process in step S320 being an affirmative determination indicates that the patient may be incontinent.

  Accordingly, in step S340, the determination unit 43 outputs an alert output instruction to the alert output unit 44, and proceeds to step S380. The alert output unit 44 that has received the alert output instruction from the determination unit 43 outputs the alert output instruction to the alert receiving unit 20 of the wearable sensor 10.

  As described above, a warning sound is output from the speaker 21 of the wearable sensor 10 of the patient, prompting the patient to wake up, and prompting the patient to go to the toilet and urinate. In addition, a warning sound is output when the patient's sleep is shallow, so it is easier for the patient to wake up than when a warning sound is output when the sleep is deep, and the patient's brain is already active. Awakening is also good.

On the other hand, if the determination in step S320 is negative, the process proceeds to step S330. In this case, for any measurement item i included in the feature amount list S2, the change tendency of the current feature amount C _i does not match the change trend of the feature amount related to the measurement item i included in the feature amount list S2, or , so that the current feature amount C _i does not exceed the threshold. Therefore, it can be estimated that there is no indication that the patient will be incontinent.

  Therefore, in step S330, the determination unit 43 determines whether or not the current time acquired in step S300 has passed the scheduled wake-up time of the patient. If the determination is affirmative, it is considered that the patient has woken up, and thereafter, there is no need to output a warning prompting the patient to urinate, and the process proceeds to step S380.

In the case of negative determination, the patient still proceeds to step S300 because it is considered to be sleeping, to newly calculate the present feature quantity C _i of the patient to repeat the steps S300 and subsequent steps.

  On the other hand, if the determination process in step S310 is affirmative, the process proceeds to step S350.

In step S350, which is the transition destination when it is determined that the sleep depth of the patient is deep, the determination unit 43 acquires the feature amount list S1 of the patient from the feature amount list DB 49. Then, the determination unit 43, among the current feature quantity C _i of patients calculated in step S300, the current and the feature amount C _i about measurement items i included in the feature list S1, acquired, included in the feature list S1 Measurement The threshold value of the feature amount related to the item i is compared for each measurement item i. Then, the determination unit 43, for all the measurement item i included in the feature list S1, the change trend of the current feature quantity C _i matches a feature amount of change trends in the measurement item i included in the feature list S1 In addition, it is determined whether or not the current feature amount C _i exceeds a threshold value. If the determination in step S350 is affirmative, the process proceeds to step S360.

  The feature amount list S1 indicates a change tendency of the feature amount that appears in common in the patient's body as a sign when the patient is incontinent when the patient's sleep depth is deep. Therefore, the determination process in step S350 being an affirmative determination indicates that the patient may be incontinent.

  Accordingly, in step S360, the determination unit 43 outputs an alert output instruction to the alert output unit 44, and proceeds to step S380. The alert output unit 44 that has received the alert output instruction from the determination unit 43 outputs the alert output instruction to the alert receiving unit 20 of the wearable sensor 10 and the alert receiving unit 36 of the information terminal 30.

  As described above, a warning sound is output from the speaker 21 of the wearable sensor 10 and the speaker 35 of the information terminal 30, and the illumination of the patient's house is turned on by the illumination control unit 37.

  For example, the information terminal 30 is installed in a room where a patient's guardian is sleeping, and the lighting control unit 37 controls lighting installed in at least one of the guardian's room and the patient's sleeping room. In this case, in addition to the patient, the patient's guardian will be further woken up. This is because it urges the patient to wake up when the patient's sleep is deep, and it is assumed that it is difficult for the patient to wake up alone compared to the case where a warning sound or the like is output when the patient's sleep is shallow. Because. If the patient's sleep is deep, the patient's guardian is also woken up in response to the detection of a sign of the patient's incontinence, and the guardian prompts the patient to go to the toilet to further increase the patient's incontinence. Can be suppressed.

  The determination unit 43 associates the time at which the alert output instruction is output to the alert output unit 44 with the patient ID as a warning time, and stores it in a predetermined area of the memory 104, for example.

On the other hand, if the determination in step S350 is negative, the process proceeds to step S370. In this case, for any measurement item i included in the feature list S1, the change trend of the current feature quantity C _i is either not match the change trend of the feature amount relating to the measurement item i included in the feature list S1, or Therefore, there is a situation in which the current feature amount C _i does not exceed the threshold value. Therefore, it can be estimated that there is no indication that the patient will be incontinent.

  Accordingly, in step S370, the determination unit 43 determines whether or not the current time acquired in step S300 has passed the scheduled wake-up time of the patient. If the determination is affirmative, it is considered that the patient has woken up, and thereafter, there is no need to output a warning prompting the patient to urinate, and the process proceeds to step S380.

In the case of negative determination, the patient still proceeds to step S300 because it is considered to be sleeping, to newly calculate the present feature quantity C _i of the patient to repeat the steps S300 and subsequent steps. The alert notification process shown in FIG.

  In step S380, the determination unit 43 stores a warning output record in which a patient ID, a warning time, and information for identifying the feature amount list S used in detecting a sign of incontinence of the patient are associated with each other, for example, in a memory 104 is stored in a predetermined area.

  Note that the warning output record is stored in the memory 104 in time series for each sleeping period of the patient. The sleeping period is, for example, a period from the scheduled sleeping time of the patient to the scheduled rising time, and the scheduled sleeping time and the scheduled rising time of the patient are input from the information terminal 30, for example.

  If no sign of incontinence of the patient is detected during the bedtime of the patient, the determination unit 43 sets the warning time to “*”, for example, and records that no warning output instruction has been issued.

  Further, the information for identifying the feature quantity list S is hereinafter referred to as “feature quantity list ID”, and the feature quantity list ID is represented by, for example, S1 and S2. When the feature amount list ID = S1, the determination unit 43 indicates that the sign of incontinence of the patient has been determined using the feature amount list S1. Further, when the feature amount list ID = S2, the determination unit 43 indicates that the sign of incontinence of the patient has been determined using the feature amount list S2.

  The alert notification process shown in FIG.

  In the above, as an example of the specific feature amount, the feature amount for each measurement item i is divided into patterns according to the sleep depth, but the pattern division method is not limited to this. For example, the feature quantity for each measurement item i may be divided into patterns according to the values of other biological data such as heart rate. Further, for example, according to the values of a plurality of biological data such as sleep depth and heart rate. Thus, the feature amount for each measurement item i may be divided into patterns.

Further, the determination unit 43 obtains the current feature value C _i and the threshold value of the feature value related to the measurement item i for all measurement items i included in the feature value list S in step S320 and step S350 of FIG. Although the comparison is made, the comparison method is not limited to this.

For example, among all measurement items i included in the feature value list S, for a predetermined number of measurement items i, the current feature value C _i and the threshold value of the feature value related to the measurement item i may be compared. good.

  As described above, in the first embodiment, the incontinence suppressing device 40 is based on the patient's biological data in a predetermined period, and the feature amount list indicating the change tendency of the feature amount that appears commonly in the patient's body as a sign that the patient is incontinent. Are generated according to the patient's sleep depth. Furthermore, the incontinence suppressing device 40 detects a sign of incontinence of the patient based on the feature amount list S corresponding to the patient's sleep depth, and changes the output destination of the warning according to the patient's sleep depth. Suppress incontinence.

  As described above, since the change tendency of the feature value differs depending on the patient's condition, specifically the patient's sleep depth, the patient's condition is predicted to be inferior compared with the case where the feature quantity list S that does not consider the patient's condition is used. It can be detected accurately.

  In addition, according to the first embodiment, as a feature amount, each patient can be obtained by using a difference between the value of the patient's biometric data on the day of incontinence and the value of the patient's biometric data on the day without incontinence. It is possible to remove the tendency of the biological information unique to the patient and extract the tendency of the biological information that appears in the patient at the time of incontinence.

  In addition, according to the first embodiment, a sign of incontinence of a patient is detected based on a feature amount list that is a combination of feature amount change trends related to a plurality of measurement items i. Therefore, a sign of incontinence of the patient can be detected with higher accuracy than in the case of using a feature amount list including a feature amount related to one measurement item i.

(Second Embodiment)
Regarding the warning output, if the incontinence suppressing device 40 instructs the warning output too early with respect to the patient's incontinence time and wakes up the patient, the patient may not yet have urine at that time. The patient may go to bed again. Therefore, even if the patient is woken up by instructing to output a warning, the patient may be incontinent.

  On the other hand, when the incontinence suppressing device 40 outputs a warning later than the time when the patient was incontinent, it is natural that the incontinence of the patient cannot be suppressed.

  In other words, if the incontinence suppressing device 40 can instruct the output of a warning during a period in which the patient feels urine that does not lead to incontinence (optimum warning period), the incontinence of the patient can be further suppressed.

  However, the time from detection of a sign of incontinence to actual incontinence may vary depending on, for example, the patient's daily physical condition and the patient's growth status.

  Therefore, in the second embodiment, it is possible to adjust the feature value threshold for each measurement item i included in the feature value list S based on the patient's incontinence time and the warning time, and to output a warning to the patient. An incontinence suppressing device that adjusts to be included in the optimum warning period as much as possible will be described.

  FIG. 8 is a diagram illustrating an example of the configuration of the incontinence suppression system 2 according to the second embodiment. The incontinence suppression system 2 is different from the incontinence suppression system 1 according to the first embodiment in that the incontinence suppression device 40 is replaced with an incontinence suppression device 40A.

  The incontinence suppressing device 40A is different from the incontinence suppressing device 40 in that an adjustment unit 46 is added.

  The adjustment unit 46 is connected to the determination unit 43 and the storage unit 45, acquires the warning time for the patient from the determination unit 43, and acquires the incontinence time of the patient from the sensor data DB 48 of the storage unit 45. Then, the adjustment unit 46 adjusts the threshold value of the feature amount included in the feature amount list S used for the determination of the incontinence sign of the patient based on the acquired incontinence time and the warning time, and is included in the feature amount list DB 49. The feature amount list S of the patient to be updated is updated with the feature amount list S with the adjusted threshold value.

  Next, FIG. 9 shows an example of a configuration diagram when the incontinence suppressing device 40A is realized by a computer.

  The configuration of the computer 100A shown in FIG. 9 is different from the configuration of the computer 100 shown in FIG. 4 in that an adjustment process 130 is added. In addition, with the addition of the adjustment process 130, the incontinence suppression program 120 according to the first embodiment is replaced with an incontinence suppression program 120A.

  The CPU 102 reads out the incontinence suppression program 120A from the storage device 106, expands it in the memory 104, and executes each process included in the incontinence suppression program 120A.

  The CPU 102 reads out the incontinence suppression program 120A from the storage device 106, expands it in the memory 104, and executes the incontinence suppression program 120A, whereby the computer 100A operates as the incontinence suppression device 40A shown in FIG.

  Further, when the CPU 102 executes the adjustment process 130, the computer 100A operates as the adjustment unit 46 illustrated in FIG.

  Note that the computer 100A can be realized by, for example, a semiconductor integrated circuit, more specifically, an ASIC or the like.

  Next, the operation of the incontinence suppressing device 40A according to the second embodiment will be described with reference to FIG. The incontinence suppressing device 40A executes, for example, a feature amount list update process for updating the feature amount list S associated with the patient at a predetermined time of one day. Here, as an example, an example of updating the feature amount list S associated with the patient instructed by the information terminal 30 will be described. However, the present invention is not limited to this, and for example, the feature amount list S is generated at a predetermined time of the day. The feature amount list update process may be executed for all the associated patients.

  Note that the feature quantity list update process updates the feature quantity list S used to determine a sign of incontinence of the patient. Therefore, the feature quantity list update process is performed by the patient from the scheduled wake-up time to the scheduled bedtime. It is preferable to execute during the time period. Further, the feature quantity list update process may be executed every day, or may be executed every week, for example, and the execution interval of the feature quantity list update process is not limited.

  First, in step S400, the adjustment unit 46 searches the sensor data DB 48 for incontinence detection data in the latest bedtime period of the patient among the incontinence detection data of the patient. In the following, the above bedtime is referred to as “previous bedtime”.

  Then, the adjustment unit 46 determines whether or not there is incontinence detection data indicating that the patient has incontinence within the previous sleeping period of the patient. If the determination is negative, the process proceeds to step S410. At this time, the adjustment unit 46 acquires the incontinence time of the patient from the incontinence detection data and stores it in a predetermined area of the memory 104.

  In step S410, the adjustment unit 46 determines whether or not a warning is output within the previous sleeping period of the patient. This determination is performed, for example, by referring to the patient warning output record stored in the memory 104 in step S380 in FIG.

  If the determination process in step S410 is negative, that is, if the patient does not incontinence and does not output a warning within the previous sleeping period, this indicates that the patient's incontinence symptoms are improving. . Accordingly, it is only necessary to observe the incontinence symptoms of the patient as it is, and since it is not necessary to update the current feature amount list S, the feature amount list update process shown in FIG.

  If the determination process in step S410 is affirmative, the process proceeds to step S420. In this case, a warning is output to the patient, but the patient is not incontinent. This situation is presumed to indicate that the patient's growth has slowed the incontinence time, since incontinence time generally slows down as the patient grows and eventually the incontinence symptoms often heal .

  Therefore, in step S420, the adjustment unit 46 determines whether or not the threshold value of the patient feature list S (used feature list) used for the determination of the incontinence sign in the previous sleeping period has been increased within a predetermined period. judge. Here, raising the threshold value of the used feature value list means increasing the absolute value of the feature value threshold value for each measurement item i included in the used feature value list.

  In the case of negative determination, the process proceeds to step S430. In step S430, the adjustment unit 46 detects a sign of the patient's incontinence later than the conventional state in accordance with the delay of the patient's incontinence time. Increase the threshold.

  In addition, when raising the threshold value of the used feature amount list, the threshold value is raised according to a predetermined variation rate. For example, as the predetermined variation rate, a plurality of variation rates having different values such as a standard variation rate (5%), a low variation rate (1%), and a high variation rate (10%) are stored in the memory 104 in advance. In step S430, the adjustment unit 46 increases the threshold value of the used feature amount list according to the low variation rate. The predetermined number of classifications of the fluctuation rates and the values of the respective fluctuation rates are examples, and are not limited thereto.

  On the other hand, if the determination process in step S420 is affirmative, the feature amount list update process shown in FIG. 10 ends.

  In this way, the adjustment unit 46 increases the threshold value of the used feature value list when the threshold value of the used feature value list has not been increased within a predetermined period. The reason for this is that the patient's growth does not progress greatly in one day, so it is considered that there is no significant change in the patient's incontinence time, for example, yesterday and today. Therefore, although a warning is output to the patient, the threshold of the patient's usage feature amount list can be adjusted at intervals of a predetermined period such as one month without adjusting the threshold every time the patient is not incontinent. This is because it is considered sufficient. For the same reason, the threshold value of the use feature amount list of the patient is adjusted according to the low variation rate among the plurality of variation rates.

  On the other hand, if the determination process in step S400 is affirmative, that is, if the patient is incontinent during the previous sleeping period, the process proceeds to step S440.

  In step S440, as in step S410, the adjustment unit 46 determines whether a warning is output during the previous sleeping period of the patient. If a negative determination is made, that is, if the patient is incontinent without a warning being output within the previous bedtime, the process proceeds to step S450.

  In step S450, the adjustment unit 46 determines whether or not the patient is incontinent before a time (pre-wake-up designated time) that is a predetermined time before the scheduled wake-up time of the patient. The predetermined time used for setting the designated time before waking up is, for example, the time required for the patient to get up and the patient sleep gradually becoming shallower. Here, 60 minutes is taken as an example, but the time is not limited to this.

  If the determination process in step S450 is negative, that is, if the patient has been incontinent after the designated time before waking up, it can be considered that the patient has been incontinent immediately before waking up. Since the incontinence time tends to be delayed as the patient's incontinence symptoms improve, this indicates that the patient's incontinence symptoms are about to heal. Accordingly, it is only necessary to observe the incontinence symptoms of the patient as it is, and since it is not necessary to update the current feature amount list S, the feature amount list update process shown in FIG.

  On the other hand, if the determination process in step S450 is affirmative, that is, if the patient has been incontinent before the designated time before waking up without a warning being output during the previous sleeping period, the process proceeds to step S460. .

In this case, the threshold value of the used feature value list is too large, and the current feature value C _i related to each measurement item i included in the used feature value list of the patient in step S320 or S350 in FIG. 7 exceeds the threshold value. It is probable that a warning was not output without being able to.

  Accordingly, in step S460, the adjustment unit 46 lowers the threshold value of the patient use feature amount list. Here, lowering the threshold value of the used feature value list means reducing the absolute value of the feature value threshold value for each measurement item i included in the used feature value list.

  At this time, the adjustment unit 46 may change the threshold reduction range depending on whether the used feature amount list is the feature amount list S1 or the feature amount list S2.

  When the patient's sleep is shallow, the time from when the warning is output until the patient wakes up is often shorter than when the patient is deeply sleep. Therefore, since the warning time when the sleep depth is shallow may be later than the warning time when the sleep depth is deep, the threshold value reduction range of the feature amount list S2 when the sleep is shallow is a feature when the sleep is deep. The threshold value is made smaller than the threshold value of the quantity list S1.

  On the other hand, if the determination process in step S440 is affirmative, that is, if the patient has been incontinent even though a warning is output within the previous sleeping period, the process proceeds to step S470.

  In step S470, the adjustment unit 46 compares the warning time for the patient with the incontinence time of the patient, and determines whether or not the incontinence time is after the warning time. In the case of a negative determination, that is, if the patient is incontinence before the warning is output, it is necessary to detect a sign of incontinence earlier than the present so that the warning is output before the patient is incontinent. The process proceeds to step S460. And as already demonstrated, in step S460, the adjustment part 46 lowers | hangs the threshold value of a patient's use feature-value list | wrist.

  On the other hand, if the determination process in step S470 is affirmative, that is, if the patient is incontinent after the warning is output, the process proceeds to step S480.

  In this case, since the warning is output before the patient is incontinent, the incontinence suppressing device 40A detects the sign of the patient's incontinence. However, it can be said that the shorter the time from the warning time to the patient's incontinence time, the more accurately the sign of the patient's incontinence is detected. Therefore, when the time from the warning time to the patient's incontinence time is longer than a predetermined time (allowable time), it is better to delay the warning time from the current state. The allowable time is stored in advance in a predetermined area of the memory 104, for example, and 30 minutes is used as an example of the allowable time. The value of the allowable time is not limited to this.

  Accordingly, in step S480, the adjustment unit 46 determines whether or not the patient's incontinence time is included within the time that has passed the allowable time from the warning time. In the case of an affirmative determination, since the warning time for the patient's incontinence time is appropriate, the feature amount list update process shown in FIG. 10 is terminated.

  On the other hand, in the case of negative determination, the process proceeds to step S490, and in step S490, the adjustment unit 46 increases the threshold value of the patient use feature amount list to make the warning time later than the current time, and the patient incontinence time is Adjust so that it is within the allowable time from the time.

  Thus, the feature quantity list update process shown in FIG. 10 is completed.

  As described above, according to the second embodiment, the incontinence suppressing device 40A adjusts the threshold value of the feature amount related to each measurement item i included in the feature amount list S based on the incontinence time and the warning time of the patient, and The time for outputting the warning is included in the optimum warning period as much as possible. Therefore, incontinence of the patient can be suppressed with higher accuracy than when the threshold value of the feature amount list S is not changed according to the relationship between the incontinence time of the patient and the warning time.

  Note that the incontinence data is not limited to being detected by the urine sensor 12, and for example, incontinence data input via the input unit 33 of the information terminal 30 may be used. In this case, the urine sensor 12 can be omitted from the wearable sensor 10.

  As described above, the disclosed technology has been described using each embodiment, but the disclosed technology is not limited to the scope described in each embodiment. Various modifications or improvements can be added to each embodiment without departing from the spirit of the disclosed technology, and forms to which the modifications or improvements are added are also included in the technical scope of the disclosed technology. For example, the processing order may be changed without departing from the scope of the disclosed technology.

  Moreover, although each embodiment demonstrated the aspect by which the incontinence suppression program 120,120A was previously memorize | stored (installed) in the memory | storage device, it is not limited to this. The incontinence suppression programs 120 and 120A according to the disclosed technology can also be provided in a form recorded on a computer-readable recording medium 116. For example, the incontinence suppression programs 120 and 120A according to the disclosed technology can be provided in a form recorded in a portable recording medium such as a CD-ROM, a DVD-ROM, and a USB memory. Further, the incontinence suppressing programs 120 and 120A according to the disclosed technology can be provided in a form recorded in a semiconductor memory such as a flash memory.

  Regarding the embodiments including the above embodiments, the following additional notes are disclosed.

(Appendix 1)
On the computer,
Based on a plurality of feature quantities indicating a patient's life activity, a characteristic showing a change tendency common when the patient is incontinent for each different state of the patient that can be distinguished by a specific feature quantity among the plurality of feature quantities Extract the quantity pattern,
Predicting incontinence of the patient based on the patient's condition and the change tendency of the feature amount indicated by the pattern according to the condition,
An incontinence suppressing method for executing processing including outputting a warning when incontinence of the patient is predicted.

(Appendix 2)
The incontinence suppressing method according to claim 1, wherein a sleep depth is used as the state of the patient, and the pattern is extracted for each different sleep depth of the patient.

(Appendix 3)
The incontinence suppressing method according to Supplementary Note 1 or Supplementary Note 2, wherein the pattern includes a plurality of feature amounts indicating a common change tendency when the patient is incontinent.

(Appendix 4)
The incontinence suppression method according to any one of supplementary notes 1 to 3, wherein a difference between the biological information of the patient when the patient is incontinent and the biological information of the patient when the patient is not incontinent is used as the feature amount.

(Appendix 5)
The output destination of the warning is changed between when the incontinence is predicted when the patient's sleep depth is shallow and when the incontinence is predicted when the patient's sleep depth is deep. Incontinence suppression method according to item.

(Appendix 6)
A threshold value corresponding to the degree of change tendency of the feature value included in the pattern is set to the feature value included in each of the extracted patterns,
Of the plurality of feature amounts according to the patient's state, the feature is predicted to be incontinent when the feature amount included in the pattern according to the patient's state exceeds the threshold value. The incontinence suppression method described.

(Appendix 7)
The incontinence suppressing method according to claim 6, wherein the threshold value is adjusted based on whether or not the warning is output and whether or not the patient is incontinent.

(Appendix 8)
The incontinence suppressing method according to claim 7, wherein the absolute value of the threshold value is reduced when the patient is incontinent before outputting the warning.

(Appendix 9)
If the patient is incontinent before outputting the warning, and the incontinence time of the patient is included in a period from when the patient goes to bed until a predetermined time, the absolute value of the threshold value is decreased. The incontinence suppression method according to appendix 7 or appendix 8.

(Appendix 10)
The absolute value of the threshold value is increased when the patient is incontinent after the warning is output, and when the period from the output of the warning to the incontinence of the patient exceeds a predetermined period. The incontinence suppression method according to any one of?

(Appendix 11)
The absolute value of the threshold is increased when the patient has not incontinence in spite of outputting the warning and the absolute value of the threshold has not been increased over a predetermined period. The incontinence suppression method according to any one of appendix 7 to appendix 10.

(Appendix 12)
Based on a plurality of feature quantities indicating a patient's life activity, a characteristic showing a change tendency common when the patient is incontinent for each different state of the patient that can be distinguished by a specific feature quantity among the plurality of feature quantities An extractor for extracting a quantity pattern;
A determination unit that predicts incontinence of the patient based on the patient's state and a change tendency of the feature amount indicated by the pattern according to the state extracted by the extraction unit;
When the determination unit predicts incontinence of the patient, an output unit that outputs a warning;
Incontinence suppression device including.

(Appendix 13)
The incontinence suppressing device according to claim 12, wherein the extraction unit uses a sleep depth as the state of the patient and extracts the pattern for each different sleep depth of the patient.

(Appendix 14)
The incontinence suppressing device according to Supplementary Note 12 or Supplementary Note 13, wherein the pattern includes a plurality of feature amounts indicating a common change tendency when the patient is incontinent.

(Appendix 15)
The extraction unit uses a difference between the patient's biological information when the patient is incontinent and the patient's biological information when the patient is not incontinent as the feature amount. Incontinence suppression device.

(Appendix 16)
The determination unit changes the output destination of the warning between when the incontinence is predicted when the sleep depth of the patient is shallow and when the incontinence is predicted when the sleep depth of the patient is deep. The incontinence suppressing device according to any one of 15.

(Appendix 17)
The extraction unit sets a threshold value corresponding to the degree of change tendency of the feature amount included in the pattern to the feature amount included in each of the extracted patterns,
The determination unit predicts that the patient is incontinent when a feature amount included in the pattern according to the patient's state exceeds the threshold among the plurality of feature amounts according to the patient's state. The incontinence suppressing device according to appendix 12 to appendix 16.

(Appendix 18)
The incontinence suppressing device according to claim 17, further comprising an adjustment unit that adjusts the threshold based on whether or not the warning is output and whether or not the patient is incontinent.

(Appendix 19)
The incontinence suppressing device according to claim 18, wherein the adjustment unit decreases the absolute value of the threshold when the patient is incontinent before the output unit outputs the warning.

(Appendix 20)
The adjusting unit includes a case where the patient is incontinent before the warning is output by the output unit, and the incontinence time of the patient is included in a period from when the patient goes to bed until a predetermined time. The incontinence suppressing device according to appendix 18 or appendix 19, wherein the absolute value of the threshold value is reduced.

(Appendix 21)
The adjusting unit is when the patient is incontinent after the output of the warning by the output unit, and when the period from the output of the warning until the patient is incontinent exceeds a predetermined period, The incontinence suppressing device according to any one of appendix 18 to appendix 20, wherein the absolute value of the threshold is increased.

(Appendix 22)
The absolute value of the threshold value when the patient does not incontinence despite the output unit outputting the warning and the absolute value of the threshold value is not increased over a predetermined period. The incontinence suppressing device according to any one of appendix 18 to appendix 21, wherein the value is increased.

(Appendix 23)
On the computer,
Based on a plurality of feature quantities indicating a patient's life activity, a characteristic showing a change tendency common when the patient is incontinent for each different state of the patient that can be distinguished by a specific feature quantity among the plurality of feature quantities Extract the quantity pattern,
Predicting incontinence of the patient based on the patient's condition and the change tendency of the feature amount indicated by the pattern according to the condition
An incontinence suppression program for executing processing including outputting a warning when incontinence of the patient is predicted.

(Appendix 24)
24. The incontinence suppression program according to appendix 23, wherein a sleep depth is used as the state of the patient, and the pattern is extracted for each different sleep depth of the patient.

(Appendix 25)
25. The incontinence suppression program according to claim 23 or 24, wherein the pattern includes a plurality of feature amounts that indicate a common change tendency when the patient is incontinent.

(Appendix 26)
26. The incontinence suppression program according to any one of appendices 23 to 25, wherein a difference between the patient's biological information when the patient is incontinent and the patient's biological information when the patient does not incontinence is used as the feature amount.

(Appendix 27)
The output destination of the warning is changed between the case where incontinence is predicted in a state where the sleep depth of the patient is shallow and the case where incontinence is predicted in a state where the sleep depth of the patient is deep. Incontinence suppression program described in the section.

(Appendix 28)
A threshold value corresponding to the degree of change tendency of the feature value included in the pattern is set to the feature value included in each of the extracted patterns,
Of the plurality of feature amounts according to the patient state, the feature amount is predicted to be incontinent when the feature amount included in the pattern according to the patient state exceeds the threshold value. The incontinence suppression program described.

(Appendix 29)
29. The incontinence suppression program according to appendix 28, wherein the threshold is adjusted based on whether the warning is output and whether the patient is incontinent.

(Appendix 30)
The incontinence suppressing program according to claim 29, wherein the absolute value of the threshold value is reduced when the patient is incontinent before outputting the warning.

(Appendix 31)
If the patient is incontinent before outputting the warning, and the incontinence time of the patient is included in a period from when the patient goes to bed until a predetermined time, the absolute value of the threshold value is decreased. The incontinence suppression program according to appendix 29 or appendix 30.

(Appendix 32)
The absolute value of the threshold value is increased when the patient is incontinent after the warning is output, and when the period from the output of the warning until the patient is incontinent exceeds a predetermined period. The incontinence suppressing program according to any one of to 31.

(Appendix 33)
The absolute value of the threshold is increased when the patient has not incontinence in spite of outputting the warning and the absolute value of the threshold has not been increased over a predetermined period. The incontinence suppressing program according to any one of appendix 29 to appendix 32.

(Appendix 34)
On the computer,
Based on a plurality of feature quantities indicating a patient's life activity, a characteristic showing a change tendency common when the patient is incontinent for each different state of the patient that can be distinguished by a specific feature quantity among the plurality of feature quantities Extract the quantity pattern,
Predicting incontinence of the patient based on the patient's condition and the change tendency of the feature amount indicated by the pattern according to the condition
A computer-readable recording medium recording an incontinence suppression program for executing processing including outputting a warning when incontinence of the patient is predicted.

DESCRIPTION OF SYMBOLS 1 (2) ... Incontinence suppression system, 10 ... Wearable sensor, 11 ... Transmission part, 12 ... Urine sensor, 13 ... Acceleration sensor, 14 ... Temperature sensor, 15 ... Respiration sensor, 16 ... heart rate sensor, 17 ... gyro sensor, 18 ... brain wave sensor, 19 ... ECG sensor, 20 ... alert receiver, 21 (35) ... speaker, 30 .. Information terminal 31... Sensor data receiving unit 32... Sensor data transmitting unit 33... Input unit 34. -Lighting control part, 40 (40A) ... Incontinence suppression device, 41 ... Collection recording part, 42 ... Extraction part, 43 ... Determination part, 44 ... Alert output part, 45 ... Storage unit, 46 ... adjustment unit, 48 ... sensor data D 49 ... feature amount list DB, S (S1, S2) ... feature amount list, 100 (100A) ... computer, 102 ... CPU, 104 ... memory, 106 ... storage device 108 ... Bus 112 ... Input device 114 ... Communication device 116 ... Recording medium 120 (120A) ... Incontinence suppression program

Claims (12)

On the computer,
Based on a plurality of feature quantities indicating a patient's life activity, a characteristic showing a change tendency common when the patient is incontinent for each different state of the patient that can be distinguished by a specific feature quantity among the plurality of feature quantities Extract the quantity pattern,
Predicting incontinence of the patient based on the patient's condition and the change tendency of the feature amount indicated by the pattern according to the condition,
An incontinence suppressing method for executing processing including outputting a warning when incontinence of the patient is predicted. The incontinence suppression method according to claim 1, wherein a sleep depth is used as the state of the patient, and the pattern is extracted for each different sleep depth of the patient. The incontinence suppressing method according to claim 1, wherein the pattern includes a plurality of feature amounts that show a common change tendency when the patient is incontinent. The incontinence suppression according to claim 1, wherein a difference between the patient's biological information when the patient is incontinent and the patient's biological information when the patient is not incontinent is used as the feature amount. Method. The output destination of the warning is changed between when the incontinence is predicted when the sleep depth of the patient is shallow and when the incontinence is predicted when the sleep depth of the patient is deep. The incontinence suppression method according to claim 1. A threshold value corresponding to the degree of change tendency of the feature value included in the pattern is set to the feature value included in each of the extracted patterns,
The patient is predicted to be incontinent when a feature amount included in the pattern according to the patient state exceeds the threshold value among the plurality of feature amounts according to the patient state. Item 6. The incontinence suppression method according to Item 5. The incontinence suppression method according to claim 6, wherein the threshold value is adjusted based on whether or not the warning is output and whether or not the patient has incontinence. The incontinence suppression method according to claim 7, wherein an absolute value of the threshold value is reduced when the patient is incontinent before outputting the warning. If the patient is incontinent before outputting the warning, and the incontinence time of the patient is included in a period from when the patient goes to bed until a predetermined time, the absolute value of the threshold value is decreased. The incontinence suppressing method according to claim 7 or 8. The absolute value of the threshold value is increased when the patient is incontinent after the warning is output, and when the period from the output of the warning to the incontinence of the patient exceeds a predetermined period. The incontinence suppression method according to any one of claims 7 to 9. When the patient has not incontinence for a predetermined period despite the output of the warning, and the absolute value of the threshold has not been increased for a predetermined period, the threshold The method for suppressing incontinence according to any one of claims 7 to 10, wherein the absolute value of is increased. Based on a plurality of feature quantities indicating a patient's life activity, a characteristic showing a change tendency common when the patient is incontinent for each different state of the patient that can be distinguished by a specific feature quantity among the plurality of feature quantities An extractor for extracting a quantity pattern;
A determination unit that predicts incontinence of the patient based on the patient's state and a change tendency of the feature amount indicated by the pattern according to the state extracted by the extraction unit;
When the determination unit predicts incontinence of the patient, an output unit that outputs a warning;
Incontinence suppression device including.

Images