Data processing apparatus, care support system, and data processing method

Abstract

To provide a data processor, a care support system, and a data processing method that can perform more appropriate processing.SOLUTION: A data processor comprises an acquisition portion capable of acquiring the first data and a processing portion capable of conducting the first operation on the basis of the first data acquired by the acquisition portion. At least a part of the first data is obtained from a first sensor capable of detecting a state of an object person. The processing portion can output first event prediction information on an occurrence of a first event related to the object person and state prediction information on a state to be predicted of the object person at the occurrence of the first event, if the first data includes first data information, in the first operation. The occurrence of the first event can be detected by a second sensor capable of detecting the state of the object person. The state to be predicted of the object person at the occurrence of the first event is not detected by the second sensor.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] Embodiments of the present invention relate to a data processing apparatus, a care support system, and a data processing method.

Background Art

[0002] In caregiving, data obtained from various sensors is used. By appropriately processing the data, more appropriate care can be provided.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Embodiments of the present invention provide a data processing apparatus, a care support system, and a data processing method capable of more appropriate processing.

Means for Solving the Problems

[0005] According to an embodiment of the present invention, a data processing apparatus includes an acquisition unit capable of acquiring first data, and a processing unit capable of performing a first operation based on the first data acquired by the acquisition unit. At least a part of the first data is obtained from a first sensor capable of detecting the state of a target person. In the first operation, when the first data includes first data information, the processing unit can output first event prediction information regarding the occurrence of a first event regarding the target person, and state prediction information regarding the predicted state of the target person at the time of the occurrence of the first event. The occurrence of the first event can be detected by a second sensor capable of detecting the state of the target person. The predicted state of the target person at the time of the occurrence of the first event is not detected by the second sensor.

Brief Description of the Drawings

[0006]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

BEST MODE FOR CARRYING OUT THE INVENTION

[0007] Hereinafter, each embodiment of the present invention will be described with reference to the drawings. In the present specification and each figure, elements that are the same as those described above with respect to the previously presented figures are denoted by the same reference numerals, and detailed descriptions thereof are omitted as appropriate.

[0008] (First Embodiment) FIG. 1 is a schematic diagram illustrating a data processing apparatus according to the first embodiment. As shown in FIG. 1, a data processing apparatus 70 according to the embodiment includes an acquisition unit 72 and a processing unit 71. A care support system 110 according to the embodiment includes the data processing apparatus 70.

[0009] The acquisition unit 72 is capable of acquiring first data DA1. The acquisition unit 72 is, for example, an interface. The processing unit 71 is capable of performing a first operation OP1 based on the first data DA1 acquired by the acquisition unit 72.

[0010] At least a part of the first data DA1 is obtained from a first sensor 31. The first sensor 31 is capable of detecting the state of a target person. For example, the target person is a care recipient. The user of the care support system 110 is a caregiver. The caregiver provides care to the care recipient.

[0011] In the embodiment, the data obtained from the first sensor 31 may be supplied to and stored in a storage unit 76. The first data DA1 stored in the storage unit 76 may be supplied to the acquisition unit 72. The storage unit 76 may be provided in a server or the like. The server may be provided at any location. The acquisition unit 72 may acquire the first data DA1 by any wireless or wired method.

[0012] In the first operation OP1, when the first data DA1 includes specific first data information DI1, the processing unit 71 is capable of outputting state prediction information SP1. In the first operation OP1, when the first data DA1 includes specific first data information DI1, the processing unit 71 may also be capable of outputting first event prediction information IP1. Thus, in the first operation OP1, when the first data DA1 includes specific data, the processing unit 71 is capable of outputting output information O01 including, for example, the first event prediction information IP1 and the state prediction information SP1.

[0013] The first event prediction information IP1 relates to, for example, the occurrence of a first event regarding a subject. The state prediction information SP1 relates to, for example, the predicted state of the subject at the time of the occurrence of the first event regarding the subject.

[0014] In one example, the first event includes at least one of getting out of bed by the subject and excretion by the subject. In this case, the predicted state of the subject at the time of the occurrence of the first event includes, for example, the prediction of at least one of falling and tumbling. Falling includes, for example, falling from a standing position or a sitting position to a position lower than those (e.g., the floor). Tumbling includes, for example, a sudden movement of the body from a bed or a chair to the floor. The predicted state may include an abnormal state of the subject at the time of the occurrence of the first event.

[0015] In an embodiment, the occurrence of the first event (e.g., getting out of bed or excretion) can be detected by a sensor (a second sensor) different from the first sensor 31. The second sensor can detect the state of the subject. The second sensor is, for example, a getting-out-of-bed sensor that detects getting out of bed or an excretion sensor that detects excretion.

[0016] On the other hand, the predicted state of the subject at the time of the occurrence of the first event is not detected by the second sensor described above. For example, when the second sensor is a getting-out-of-bed sensor or an excretion sensor, the predicted state (such as a high possibility of falling or tumbling) is not detected by the second sensor.

[0017] Thus, in the first operation OP1 in the embodiment, the processing unit 71 outputs the first event prediction information IP1 regarding the prediction of the occurrence of the first event detectable by the second sensor. In addition to this, the processing unit 71 outputs additional information not detected by the second sensor as the state prediction information SP1. For example, in addition to the prediction of the occurrence of getting out of bed or excretion (the first event prediction information IP1), the prediction that the possibility of falling or tumbling at the time of the first event is higher than normal (the state prediction information SP1) is output.

[0018] The user (caregiver) of the data processing apparatus 70 according to the embodiment can know not only the prediction of the occurrence of the first event regarding the target person (for example, the care recipient), but also the prediction of the abnormal state that is likely to occur at that time. According to the embodiment, a data processing apparatus capable of more appropriate processing can be provided.

[0019] As shown in FIG. 1, for example, the output information O01 (the first event prediction information IP1 and the state prediction information SP1) from the processing unit 71 is supplied to the communication unit 75. The communication unit 75 can provide the output information O01 to an external device. The external device is, for example, the terminal device 81 used by the user. The terminal device 81 is, for example, a communication terminal or a computer. The terminal device 81 may include, for example, a portable electronic device (such as a smartphone). For example, the terminal device 81 includes an output unit 82. The output unit 82 includes, for example, a display. For example, the output information O01 (the first event prediction information IP1 and the state prediction information SP1) is displayed on the output unit 82. The output unit 82 can output the first event prediction information IP1 and the state prediction information SP1.

[0020] For example, the first event prediction information IP1 and the state prediction information SP1 are displayed on the output unit 82. The first event prediction information IP1 may include, for example, at least any one of character information, image information, and pictograms. The state prediction information SP1 may include, for example, at least any one of character information, image information, and pictograms.

[0021] For example, when the first event is getting out of bed, the output unit 82 can display, as the first event prediction information IP1, character information (text information) regarding the prediction of getting out of bed. The output unit 82 can display, as the first event prediction information IP1, character information (text information) regarding the time when getting out of bed is predicted to occur. The output unit 82 can display, as the first event prediction information IP1, image information or a pictogram indicating that getting out of bed is predicted.

[0022] For example, when the first event is getting out of bed, the output unit 82 can display, as the state prediction information SP1, character information (text information) about the prediction of an abnormal state (for example, a high possibility of falling or tumbling, etc.). The output unit 82 can display, as the state prediction information SP1, image information or a pictogram indicating that an abnormal state is predicted.

[0023] The user can obtain, through the output (for example, display) by the output unit 82, the prediction (possibility) of the occurrence of the first event of the target person and the prediction (possibility) regarding the state (such as an abnormal state) at the time of the first event. Thereby, the user can know the possibility of the occurrence of the first event before the occurrence of the first event. For example, by knowing the possibility of an abnormal state in advance, the user can provide more appropriate care to the care recipient.

[0024] The care support system 110 according to the embodiment may include the data processing device 70 and the first sensor 31. The care support system 110 may further include the terminal device 81. According to the care support system 110 according to the embodiment, more appropriate care can be provided to the care recipient. A plurality of terminal devices 81 may be provided for one data processing device 70.

[0025] The output unit 82 provided in the terminal device 81 may output the first event prediction information IP1 and the state prediction information SP1 by sound. The output unit 82 may include a speaker. The output unit 82 may output the first event prediction information IP1 and the state prediction information SP1 by vibration or the like.

[0026] In an embodiment, the state prediction information SP1 includes, for example, a prediction that the subject will rush to perform a first event. For example, as described above, when the first data DA1 includes the first data information DI1, the processing unit 71 can predict the occurrence of a first event (such as getting out of bed). At this time, the processing unit 71 can predict, based on the first data information DI1, that the subject is likely to rush to perform the first event (such as getting out of bed). If the subject rushes to perform the first event, there is a high possibility of falling or tripping. Thus, the state prediction information SP1 includes, for example, a prediction of an abnormal state.

[0027] The abnormal state includes, for example, performing the first event more hurriedly than usual as described above. The abnormal state includes, for example, falling or tripping. The abnormal state may include at least any one of incontinence, aspiration, insomnia, wandering, and physical condition abnormalities. An example of the derivation of the state prediction information SP1 will be described later.

[0028] Hereinafter, examples of sensors used in the embodiment will be described. FIG. 2 is a schematic diagram illustrating a data processing apparatus according to the first embodiment. As shown in FIG. 2, the sensor is used, for example, together with a bed 86 used by a subject 85 (such as a care recipient). The sensor may include, for example, at least any one of a body movement sensor 41, an imaging device 42, a weight sensor 43, an excretion device 44, and an excretion sensor 45.

[0029] The body movement sensor 41 is provided, for example, on the bed 86 used by the subject 85. The body movement sensor 41 is provided, for example, between the bottom of the bed 86 and the mattress. The body movement sensor 41 can detect, for example, vibrations associated with the body movement of the subject 85.

[0030] The imaging device 42 can image the subject 85. The imaging device 42 may be able to image a space including the bed 86 used by the subject 85. For example, the imaging device 42 can image the subject 85 on the bed 86. The imaging device 42 may also be able to image the subject 85 away from the bed 86.

[0031] The weight sensor 43 is provided, for example, on the floor on which the bed 86 is placed. For example, it can detect the weight from the subject 85. The subject 85 gets up from the bed 86 and moves onto the weight sensor 43. At this time, the weight sensor 43 detects the weight from the subject 85. Thereby, the weight sensor 43 can detect the getting out of bed of the subject 85. The weight sensor 43 is, for example, a getting out of bed sensor.

[0032] The excretion device 44 is a toilet used by the subject 85. The excretion device 44 can detect the excretion of the subject 85.

[0033] The excretion sensor 45 may be provided, for example, on the mattress of the bed 86. The excretion sensor 45 can detect, for example, the odor associated with excretion. The excretion sensor 45 is, for example, an odor sensor. The excretion sensor 45 can detect the excretion of the subject 85 based on, for example, the odor. In an embodiment, these sensors may be used in various combinations.

[0034] In the first example, the first sensor 31 includes the body movement sensor 41. On the other hand, the second sensor includes the excretion device 44 or the excretion sensor 45. In this case, the first event is excretion. Before the excretion of the subject 85, for example, the body movement becomes large. By detecting the magnitude of the body movement and the like, the first event (excretion) can be predicted. For example, based on the temporal change of the body movement and the like, it is possible to detect whether the subject 85 is in a hurry more than usual. Such information regarding the body movement becomes the first data DA1. In an embodiment, based on the first data DA1 obtained by the first sensor 31 (body movement sensor 41), the first event prediction information IP1 regarding the first event (excretion) and the state prediction information SP1 (abnormal states such as being in a hurry, falling or toppling) at the time of the first event are predicted. Excretion can be detected by the second sensor. The state prediction information SP1 (abnormal states such as being in a hurry, falling or toppling) is not detected by the second sensor.

[0035] In the second example, the first sensor 31 includes the imaging device 42. On the other hand, the second sensor includes the excretion device 44 or the excretion sensor 45. In this case, the first event is excretion. Information regarding the magnitude of body movement before excretion of the subject 85, and information regarding the temporal change of body movement, etc. are obtained by the imaging device 42. Also in the second example, based on the first data DA1 obtained by the first sensor 31 (imaging device 42), first event prediction information IP1 regarding the first event (excretion), and state prediction information SP1 (abnormal states such as being in a hurry, falling, or tumbling) at the time of the first event are predicted. Excretion is detectable by the second sensor. The state prediction information SP1 (abnormal states such as being in a hurry, falling, or tumbling) is not detected by the second sensor.

[0036] In the third example, the first sensor 31 includes the body movement sensor 41. On the other hand, the second sensor includes the load sensor 43. In this case, the first event is getting out of bed. Before the subject 85 gets out of bed, for example, the body movement becomes large. By detecting the magnitude of the body movement, etc., the first event (getting out of bed) can be predicted. For example, based on the temporal change of the body movement, etc., it is possible to detect whether the subject 85 is in a hurry more than usual. Such information regarding the body movement becomes the first data DA1. In the embodiment, based on the first data DA1 obtained by the first sensor 31 (body movement sensor 41), first event prediction information IP1 regarding the first event (getting out of bed), and state prediction information SP1 (abnormal states such as being in a hurry, falling, or tumbling) at the time of the first event are predicted. Getting out of bed is detectable by the second sensor. The state prediction information SP1 (abnormal states such as being in a hurry, falling, or tumbling) is not detected by the second sensor.

[0037] In the fourth example, the first sensor 31 includes the imaging device 42. On the other hand, the second sensor includes the load sensor 43. In this case, the first event is getting out of bed. Information regarding the magnitude of body movement before the subject 85 gets out of bed, and information regarding temporal changes in body movement, etc., are obtained by the imaging device 42. Based on the first data DA1 obtained by the first sensor 31 (imaging device 42), first event prediction information IP1 regarding the first event (getting out of bed), and state prediction information SP1 (abnormal states such as being in a hurry, falling, or toppling) at the time of the first event, are predicted. Getting out of bed can be detected by the second sensor. The state prediction information SP1 (abnormal states such as being in a hurry, falling, or toppling) is not detected by the second sensor.

[0038] Thus, in the embodiment, the first sensor 31 may include at least one of the body movement sensor 41 provided on the bed 86 used by the subject 85 and the imaging device 42 that images the subject 85. For example, the second sensor may include at least one of the excretion device 44 used by the subject 85 and the excretion sensor 45 that can detect the excretion of the subject 85.

[0039] In another example, the first sensor 31 includes at least one of the body movement sensor 41 and the imaging device 42. The second sensor includes the load sensor 43 that can detect the load from the subject 85.

[0040] The second sensor is, for example, a binary sensor that detects the presence or absence of the first event (such as getting out of bed or excretion). On the other hand, the first sensor 31 is a multi - value sensor that continuously detects the state of the subject 85. The change in the state of the subject 85 that occurs before the occurrence of the first event is detected by the first sensor 31. Based on the detection result of the change in the state of the subject 85 by the first sensor 31, the occurrence of the first event is predicted, and the state (abnormal state) of the subject 85 at the time of the first event is predicted.

[0041] The second sensor is, for example, a "poor sensor". The first sensor 31 is, for example, a "rich sensor". By combining these, the abnormal state of the subject 85 can be detected more accurately. In the embodiment, various modifications are possible for such a combination of the first sensor 31 and the second sensor. As will be described later, for example, the amount of information obtained from the first sensor 31 is larger than the amount of information obtained from the second sensor.

[0042] Hereinafter, an example of deriving the first event prediction information IP1 and the state prediction information SP1 will be described.

[0043] FIGS. 3(a) to 3(c) are schematic diagrams illustrating the operation of the data processing apparatus according to the first embodiment. The horizontal axis of these figures is the time tm. The vertical axis of FIG. 3(a) is the signal strength SS1 of the signal detected by the first sensor 31. The signal strength SS1 corresponds to, for example, the first data DA1. The vertical axis of FIG. 3(b) is the signal strength corresponding to the output information O01. FIG. 3(b) illustrates the output of the first event prediction information IP1 and the state prediction information SP1 from the processing unit 71. The vertical axis of FIG. 3(c) is the signal strength SS2 of the signal related to the occurrence of the first event detected by the second sensor.

[0044] As shown in FIG. 3(c), at the time tm2, the occurrence of the first event detected by the second sensor is detected. For example, the second sensor detects getting out of bed or excretion at the time tm2.

[0045] The first data DA1 illustrated in FIG. 3(a) is supplied to the processing unit 71. As shown in FIG. 3(a), the signal strength SS1 of the signal detected by the first sensor 31 changes with time tm. The signal strength SS1 corresponds to, for example, the magnitude of body movement. For example, a first threshold value SV1 is determined for the signal strength SS1. For example, a state ti1 in which the signal strength SS1 exceeds the first threshold value SV1 is detected by the processing unit 71. For example, the frequency tc1 of the state ti1 in which the signal strength SS1 exceeds the first threshold value SV1 may be detected by the processing unit 71. For example, the occurrence time tm1 of the state ti1 in which the signal strength SS1 exceeds the first threshold value SV1 may be detected by the processing unit 71. For example, the duration td1 of the state ti1 in which the signal strength SS1 exceeds the first threshold value SV1 may be detected by the processing unit 71. Based on these detection results, the processing unit 71 can derive first event prediction information IP1 and state prediction information SP1.

[0046] For example, when the duration td1 is longer than a second threshold value, a first event (such as excretion) can be predicted. When the duration td1 is longer than the second threshold value and the frequency tc1 is higher than a third threshold value, it can be predicted that the subject is in a hurry to perform the first event (such as excretion). Thereby, the first event prediction information IP1 and the state prediction information SP1 are derived. The state prediction information SP1 includes, for example, an abnormal state. The abnormal state includes, for example, being in a hurry, or having a higher possibility of falling or toppling than usual.

[0047] In this way, the first data information DI1 regarding the first data DA1 may include at least any one of a state ti1 in which the signal strength SS1 corresponding to the body movement of the subject 85 exceeds the first threshold value SV1, the frequency tc1 of the state ti1, the occurrence time tm1 of the state ti1, and the duration td1 of the state ti1. When the first data DA1 includes such first data information DI1, the processing unit 71 derives the first event prediction information IP1 and the state prediction information SP1 according to the first data information DI1. The occurrence time tm1 of the state ti1 is earlier than the occurrence time tm2 of the first event detected by the second sensor.

[0048] As shown in FIG. 3(b), the processing unit 71 outputs the first event prediction information IP1 and the state prediction information SP1 before the time tm2 of the occurrence of the first event detected by the second sensor.

[0049] The first event prediction information IP1 and the state prediction information SP1 are supplied to the user via the terminal device 81. The user appropriately responds based on this information.

[0050] The operation of such a processing unit 71 can be implemented by, for example, machine learning. Hereinafter, an example of the machine learning of the processing unit 71 according to the embodiment will be described.

[0051] FIG. 4 is a schematic diagram illustrating the operation of the data processing device according to the first embodiment. FIG. 4 illustrates the second operation OP2 in the data processing device 70. The second operation OP2 is an operation in the machine learning mode. At least a part of the second operation OP2 is performed before the first operation OP1.

[0052] In the second operation OP2, the processing unit 71 can acquire the first information I1 obtained from the first sensor 31 and the second information I2 obtained from the second sensor 32. These pieces of information may be supplied to the processing unit 71 via the acquisition unit 72. These pieces of information may be stored in the storage unit 76, and the information stored in the storage unit 76 may be supplied to the processing unit 71.

[0053] Based on the first information I1 and the second information I2, the processing unit 71 can derive a machine learning model 88. The processing unit 71 can perform the first operation OP1 based on the machine learning model 88. The machine learning model 88 may include various functions such as polynomials, for example.

[0054] Hereinafter, an example of the processing in the processing unit 71 in the second operation OP2 will be described. FIGS. 5(a) and 5(b) are schematic diagrams illustrating the operation of the data processing device according to the first embodiment. The horizontal axis of these figures is the time tm. The vertical axis of FIG. 5(a) is the signal intensity IS1 of the signal detected by the first sensor 31. The signal intensity IS1 corresponds to, for example, the first information I1. The vertical axis of FIG. 5(b) is the intensity IS2 of the signal related to the occurrence of the learning event detected by the second sensor 32.

[0055] The first sensor 31 is, for example, a multi-value sensor that continuously detects the state of the subject 85. The second sensor 32 is a binary sensor that detects the presence or absence of the first event (such as getting out of bed or excretion). The amount of the first information I1 is larger than the amount of the second information I2.

[0056] As shown in FIG. 5(b), the second information I2 includes learning event occurrence information IL0 related to the occurrence of the learning event. The learning event corresponds to the above-mentioned first event. The occurrence of the learning event is detected by the second sensor 32. When the first event is getting out of bed, the learning event is getting out of bed during learning. When the first event is excretion, the learning event is excretion during learning. The learning event occurrence information IL0 includes information related to detecting the occurrence of getting out of bed or excretion, etc. The occurrence of the learning event is performed, for example, at the time tL0.

[0057] As shown in FIG. 5(a), the first information I1 includes first period information Itp1. The first period information Itp1 is obtained from the first sensor 31 during the first period tp1. The first period tp1 includes the time before the occurrence of the learning event (time tL0).

[0058] The first sensor 31 is continuously detecting the state of the subject 85. The first information I1 obtained from the first sensor 31 includes information before the occurrence of the learning event (time tL0). For example, the processing unit 71 may extract information in the first period tp1 before the occurrence of the learning event (time tL0) from the continuous first information I1.

[0059] The first period information Itp1 corresponding to the first period tp1 is used as part of the teacher data in machine learning. The first period information Itp1 is, for example, an explanatory variable.

[0060] Learning event generation information IL0 regarding the generation of learning events is used as part of the teacher data in machine learning. The learning event generation information IL0 is, for example, a target variable.

[0061] In the second operation OP2, the processing unit 71 can derive the machine learning model 88 using the learning event generation information IL0 and the information based on the first period information Itp1 as teacher data.

[0062] In the first operation OP1, by inputting the first data DA1 into the machine learning model 88 derived in the second operation OP2, the occurrence of the first event can be predicted.

[0063] In the second operation OP2, the processing unit 71, for example, adds a first annotation to the first period information Itp1 based on the learning event generation information IL0. For example, the teacher data includes the first period information Itp1 with the first annotation added.

[0064] For example, the first period tp1 is a period including the time before the occurrence time tL0 of the learning event. This period may be predetermined. Regarding the first period information Itp1 in the first period tp1, a first annotation regarding the presence or absence of the learning event is added. The addition of the first annotation may be performed by the processing unit 71. By adding the annotation by the processing unit 71 included in the data processing device 70, machine learning can be efficiently and rapidly implemented.

[0065] As described above, the processing unit 71 can perform the first operation OP1 based on the first data DA1 acquired by the acquisition unit 72. As described above, at least a part of the first data DA1 is obtained from the first sensor 31. In an embodiment, a part of the first data DA1 may be obtained from another sensor (for example, the third sensor).

[0066] FIG. 6 is a schematic diagram illustrating the operation of the data processing device according to the first embodiment. As shown in FIG. 6, part of the first data DA1 may be obtained from the third sensor 33 capable of detecting the state of the subject 85.

[0067] In one example, the first sensor 31 includes a body movement sensor 41 provided on the bed used by the subject 85. The third sensor 33 includes an imaging device 42 that images the subject 85. The second sensor 32 includes at least one of an excretion device 44 used by the subject 85 and an excretion sensor 45 capable of detecting the excretion of the subject 85. The first data DA1 including the detection result by the first sensor 31 and the detection result by the third sensor 33 is input to the machine learning model 88. Thereby, the first event prediction information IP1 and the state prediction information SP1 are output. In this case, the first event includes excretion. The state at the time of occurrence of the first event includes, for example, prediction of an abnormal state (at least one of being in a hurry, falling, and tumbling).

[0068] In another example, the first sensor 31 includes a body movement sensor 41. The third sensor 33 includes an imaging device 42. The second sensor 32 includes a weight sensor 43 capable of detecting the weight from the subject 85. In this case, the first event includes getting out of bed. The state at the time of occurrence of the first event includes, for example, prediction of an abnormal state (at least one of being in a hurry, falling, and tumbling).

[0069] For example, in the second operation OP2 (operation in the machine learning mode), the information obtained from the third sensor 33 may be used.

[0070] As shown in FIG. 6, for example, in the second operation OP2, the processing unit 71 can acquire the first information I1 obtained from the first sensor 31, the second information I2 obtained from the second sensor 32, and the third information I3 obtained from the third sensor 33. In the second operation OP2, the processing unit 71 can derive the machine learning model 88 using the information including these as teacher data. The processing unit 71 can perform the first operation OP1 based on the machine learning model 88.

[0071] Figs. 7(a) to 7(c) are schematic diagrams illustrating the operations of the data processing apparatus according to the first embodiment. The horizontal axis in these figures is the time tm. The vertical axis in Fig. 7(a) is the signal intensity IS1 of the signal detected by the first sensor 31. The signal intensity IS1 corresponds to, for example, the first information I1. The vertical axis in Fig. 7(b) is the signal intensity IS3 of the signal (which may be a video signal) detected by the third sensor 33. The signal intensity IS3 corresponds to, for example, the third information I3. The vertical axis in Fig. 7(c) is the intensity IS2 of the signal related to the occurrence of the learning event detected by the second sensor 32.

[0072] The amount of the first information I1 is larger than the amount of the second information I2. In this example, the amount of the third information I3 is larger than the amount of the second information I2.

[0073] As shown in Fig. 7(c), the second information I2 includes the second event occurrence information IL2 related to the occurrence of the second event detected by the second sensor 32. The second event corresponds to the first event. The second event is an event during machine learning. The second event is detected at the time tL2.

[0074] As shown in Fig. 7(a), the first information I1 includes the first period information Itp1. The first period information Itp1 includes the information (at least a part of the first information I1) obtained from the first sensor 31 during the first period tp1 that includes the time before the occurrence of the second event (time tL2).

[0075] As shown in Fig. 7(b), the third information I3 includes the second period information Itp2. The second period information Itp2 includes the information (at least a part of the third information I3) obtained from the third sensor 33 during the second period tp2 that includes the time before the occurrence of the second event (time tL2).

[0076] In the second operation OP2, the processing unit 71 can derive the machine learning model 88 (see Fig. 6) using the second event occurrence information IL2, the information based on the first period information Itp1, and the information based on the second period information Itp2 as teacher data. The processing unit 71 can perform the first operation OP1 based on the machine learning model 88.

[0077] In the second operation OP2, the processing unit 71 may add a first annotation to the first period information Itp1 and a second annotation to the second period information Itp2 based on the second event occurrence information IL2. The training data includes, for example, the first period information Itp1 to which the first annotation is added and the second period information Itp2 to which the second annotation is added.

[0078] By using the first information I1 obtained from the first sensor 31 and the third information I3 obtained from the third sensor 33, for example, the first event prediction information IP1 can be derived more accurately. By using the first information I1 and the third information I3, for example, the state prediction information SP1 can be derived more accurately. By using the first information I1 and the third information I3, for example, the types of the state prediction information SP1 increase.

[0079] For example, by using a plurality of types of information (such as the first information I1 and the third information I3) obtained from different sensors (the first sensor 31 and the third sensor 33), an abnormal state can be derived more accurately, for example.

[0080] FIGS. 8(a) to 8(c) are schematic diagrams illustrating the operations of the data processing apparatus according to the first embodiment. The horizontal axis of these figures is the time tm. The vertical axis of FIG. 8(a) is the signal intensity IS1 of the signal detected by the first sensor 31. The signal intensity IS1 corresponds to the first information I1, for example. In the example of FIG. 8(a), the first sensor 31 includes the body motion sensor 41. The vertical axis of FIG. 8(b) is the signal intensity IS3 of the signal detected by the third sensor 33. The signal intensity IS3 corresponds to the third information I3, for example. In the example of FIG. 8(b), the third sensor 33 includes the weighted sensor 43. The vertical axis of FIG. 8(c) is the intensity IS2 of the signal related to the occurrence of the learning event detected by the second sensor 32. In the example of FIG. 8(c), the second sensor 32 includes the excretion device 44.

[0081] In the examples of FIGS. 8(a) to 8(c), the first sensor 31 detects the time (duration td1) of body movement exceeding the threshold value. The third sensor 33 can detect the time from getting out of bed until starting to walk. The second sensor 32 detects excretion. In this example, when the time from getting out of bed until starting to walk is longer than the threshold value (normal value), the processing unit 71 outputs state prediction information SP1 regarding the abnormal state. The abnormal state includes, for example, at least one of a hurry state, a fall, and a tumble.

[0082] As described above, the processing unit 71 may output the first event prediction information IP1 and the state prediction information SP1 based on the first data DA1 obtained by the plurality of sensors (the first sensor 31 and the third sensor 33).

[0083] In the examples of FIGS. 8(a) to 8(c), for example, the second period tp2 may be after the first period tp1. The second period tp2 may be later than the first period tp1. Information (at least a part of the third information I3) may be obtained from the third sensor 33 in the second period tp2 after the first period tp1 in which information (at least a part of the first information I1) is obtained from the first sensor 31.

[0084] In the embodiment, the processing unit 71 may add an annotation based on information from the plurality of sensors (the first sensor 31 and the third sensor 33). The acquisition period of the data used for adding the annotation (for example, the above-mentioned first period tp1 and second period tp2) may be changeable (specifiable).

[0085] For example, when an event is predicted based on the plurality of sensors (the first sensor 31 and the third sensor 33), the prediction period of one of the plurality of sensors and the prediction period of the other of the plurality of sensors may be different. By being able to specify different periods, more accurate prediction becomes possible.

[0086] When an event is predicted based on a plurality of sensors (the first sensor 31 and the third sensor 33), the spatial positions in the detection by the plurality of sensors may be different from each other. For example, the detection of the movement of the subject 85 (such as walking) and the state of the living room of the subject 85 (such as temperature and humidity) may be detected by a plurality of sensors. In this case, the locations of the detections by the plurality of sensors are different from each other. In such a case, the processing unit 71 may assign an annotation based on the information obtained by the plurality of sensors (the first sensor 31 and the third sensor 33). The detection location of the data used when assigning the annotation may be changed (specified). This enables more accurate prediction.

[0087] In an embodiment, a plurality of events may be predicted. For example, the acquisition unit 72 can acquire the first data DA1. The processing unit 71 may be able to perform a first operation OP1 based on the first data DA1 acquired by the acquisition unit 72. At least a part of the first data DA1 is obtained from the first sensor 31 that can detect the state of the subject 85. In the first operation OP1, when the first data DA1 includes the first data information DI1, the processing unit 71 may be able to output first event prediction information IP1 regarding the occurrence of a first event regarding the subject 85 and second event prediction information regarding the occurrence of a second event after the first event. The occurrence of the first event can be detected by the second sensor 32 that can detect the state of the subject 85. The occurrence of the second event can be detected by the third sensor 33 that can detect the state of the subject 85.

[0088] For example, the first event includes getting out of bed. The second event includes excretion. In this example, the first sensor 31 is the body motion sensor 41. Based on the data obtained from the first sensor 31, getting out of bed is predicted as the first event and excretion is predicted as the second event. For example, a plurality of chained events may be predicted.

[0089] Regarding a plurality of events, the processing unit 71 may add annotations. For example, when excretion is detected by the second sensor 32, an annotation may be added to the information obtained from the first sensor 31, and further, an annotation may be added to the information obtained from the third sensor 33. For example, based on the result, state prediction may be performed more appropriately.

[0090] For example, based on the detection result of the second sensor 32, an annotation is added to the detection results of a plurality of sensors (the first sensor 31 and the third sensor 33). For example, even when the duration of large body movement is longer than a threshold value (normal value), state prediction information SP1 regarding an abnormal state is output.

[0091] For example, the first sensor 31 includes a body movement sensor 41 provided on a bed 86 used by the subject 85. The second sensor 32 includes a weight sensor 43 capable of detecting the weight from the subject 85. The third sensor 33 includes at least one of an excretion device 44 used by the subject 85 and an excretion sensor 45 capable of detecting the excretion of the subject 85.

[0092] Figs. 9(a) to 9(d) are schematic diagrams illustrating sensors used in the data processing apparatus according to the first embodiment.

[0093] As shown in Fig. 9(a), the sensor used in the data processing apparatus 70 may include a moisture sensor 46. The moisture sensor 46 can detect the moisture ingested by the subject 85. The moisture sensor 46 may be provided, for example, in a beverage container used by the subject 85. The moisture sensor 46 can detect, for example, the weight or the height of the liquid level, and based on the result, can detect the amount of moisture ingested by the subject 85. The moisture sensor 46 may be, for example, a device for managing the moisture ingested by the subject 85.

[0094] As shown in FIG. 9(b), the sensor used in the data processing device 70 may include a food and drink sensor 47. The food and drink sensor 47 can detect the food and drink intake of the subject 85. The food and drink sensor 47 may be able to detect, for example, the amount (weight) of food consumed by the subject 85. The food and drink sensor 47 may include, for example, a weighing instrument. The food and drink sensor 47 may also include, for example, an imaging element 47a. The amount of food may be detected based on the image captured by the imaging element 47a.

[0095] As shown in FIG. 9(c), the sensor used in the data processing device 70 may include a posture sensor 48. The posture sensor 48 can detect the posture of the subject 85. The posture sensor 48 may, for example, also detect the weight borne by the subject 85. The posture sensor 48 may be able to optically detect the posture of the subject 85, for example. In this example, the posture sensor 48 is provided on the wheelchair 86A. The posture sensor 48 may be provided at any location.

[0096] As shown in FIG. 9(d), the sensor used in the data processing device 70 may include a swallowing sensor 49. The swallowing sensor 49 can detect the swallowing state of the subject 85. The swallowing sensor 49 may be able to detect, for example, the sound during the swallowing of the subject 85. Based on the detection result of the sound, the swallowing state is detected. When the detected sound is different from normal, abnormal swallowing can be determined. The swallowing sensor 49 can detect abnormal swallowing of the subject 85. The swallowing sensor 49 may be attached to, for example, the neck of the subject 85. In the swallowing sensor 49, the swallowing state may be detected by any method.

[0097] Thus, the sensor used in the data processing device 70 may include at least any one of, for example, a body movement sensor 41, an imaging device 42, a weight sensor 43, an excretion device 44, an excretion sensor 45, a moisture sensor 46, a food and drink sensor 47, a posture sensor 48, and a swallowing sensor 49.

[0098] In one example, the first event includes at least one of water intake, body movement, and excretion. The predicted state of the subject 85 at the time of occurrence of the first event includes a prediction of incontinence. The state prediction information SP1 includes a prediction of a first event (at least one of water intake, body movement, and excretion) that is different from normal. For example, the state prediction information SP1 may include information regarding at least one of water intake different from normal, body movement different from normal, and excretion different from normal. In this example, the first sensor 31 includes at least one of a body movement sensor 41 provided on the bed 86 used by the subject 85 and a water sensor 46 capable of detecting the water intake of the subject 85. For example, the second sensor 32 includes at least one of an excretion device 44 used by the subject 85 and an excretion sensor 45 capable of detecting the excretion of the subject 85.

[0099] In another example, the first event includes at least one of eating and drinking, body movement, and a change in posture. For example, the predicted state of the subject 85 at the time of occurrence of the first event includes a prediction of aspiration. The state prediction information SP1 includes a prediction of a first event (at least one of eating and drinking, body movement, and a change in posture) that is different from normal. In this example, the first sensor 31 includes at least one of a body movement sensor 41 provided on the bed used by the subject 85, an eating and drinking sensor 47 capable of detecting the eating and drinking of the subject 85, and a posture sensor 48 capable of detecting the posture of the subject 85. The second sensor 32 includes a swallowing sensor 49 capable of detecting abnormal swallowing of the subject 85.

[0100] Also in the example where the various sensors illustrated in FIGS. 9(a) to 9(d) are used, the various processes described with respect to FIGS. 3 to 8 can be applied.

[0101] FIG. 10 is a schematic diagram illustrating a data processing apparatus according to the first embodiment. As shown in FIG. 10, the data processing apparatus 70 includes, for example, a processing unit 71, an acquisition unit 72, and a storage unit 73. The processing unit 71 is, for example, an electric circuit. The storage unit 73 may include, for example, at least one of a ROM (Read Only Memory) and a RAM (Random Access Memory). Any storage device may be used as the storage unit 73.

[0102] The data processing apparatus 70 may include a display unit 79b, an input unit 79c, and the like. The display unit 79b may include various displays. The input unit 79c includes, for example, a device having an operation function (such as a keyboard, a mouse, a touch input panel, or a voice recognition input device).

[0103] The embodiment may include a program. The program causes a computer (processing unit 71) to perform the above-described operations. The embodiment may include a storage medium storing the above program.

[0104] (Second Embodiment) FIG. 11 is a flowchart illustrating a data processing method according to the second embodiment. As shown in FIG. 11, the data processing method according to the embodiment causes the processing unit 71 to obtain first data DA1 (step S110). The data processing method causes the processing unit 71 to perform a first operation OP1 based on the first data DA1. At least a part of the first data DA1 is obtained from a first sensor 31 capable of detecting the state of the subject 85.

[0105] In the first operation OP1, when the first data DA1 includes the first data information DI1, the processing unit 71 can output first event prediction information IP1 and state prediction information SP1 (step S120). The first event prediction information IP1 relates to the occurrence of a first event regarding the subject 85. The state prediction information SP1 relates to the predicted state of the subject 85 at the time of the occurrence of the first event. The occurrence of the first event can be detected by a second sensor 32 capable of detecting the state of the subject 85. The predicted state of the subject 85 at the time of the occurrence of the first event is not detected by the second sensor 32. The data processing method according to the embodiment may be capable of implementing the first operation OP1 described with respect to the first embodiment. The data processing method according to the embodiment may be capable of implementing the second operation OP2 described with respect to the first embodiment.

[0106] The embodiment may include the following configuration. (Configuration 1) An acquisition unit capable of acquiring first data, A processing unit capable of performing a first operation based on the first data acquired by the acquisition unit, Comprising, At least a part of the first data is obtained from a first sensor capable of detecting the state of the subject, In the first operation, when the first data includes first data information, the processing unit can output first event prediction information regarding the occurrence of a first event regarding the subject and state prediction information regarding the predicted state of the subject at the time of the occurrence of the first event, The occurrence of the first event can be detected by a second sensor capable of detecting the state of the subject, The predicted state of the subject at the time of the occurrence of the first event is not detected by the second sensor, a data processing device.

[0107] (Configuration 2) The first event includes at least one of getting out of bed and excretion, The predicted state of the subject at the time of the occurrence of the first event includes at least one of predicting a fall and a tumble, the data processing device according to Configuration 1.

[0108] (Configuration 3) The state prediction information includes a prediction that the subject will rush to perform the first event, and the data processing apparatus according to Configuration 2.

[0109] (Configuration 4) The first sensor includes at least one of a body movement sensor provided on the bed used by the subject and an imaging device that images the subject. The second sensor includes at least one of an excretion device used by the subject and an excretion sensor capable of detecting the excretion of the subject, and the data processing apparatus according to Configuration 2 or 3.

[0110] (Configuration 5) The first sensor includes at least one of a body movement sensor provided on the bed used by the subject and an imaging device that images the subject. The second sensor includes a weight sensor capable of detecting the weight from the subject, and the data processing apparatus according to Configuration 2 or 3.

[0111] (Configuration 6) The first event includes at least one of water intake, body movement, and excretion. The predicted state of the subject at the time of occurrence of the first event includes a prediction of incontinence, and the data processing apparatus according to Configuration 1.

[0112] (Configuration 7) The first event includes at least one of water intake, body movement, and excretion. The state prediction information includes a prediction of the first event that is different from normal, and the data processing apparatus according to Configuration 6.

[0113] (Configuration 8) The first sensor includes at least one of a body movement sensor provided on the bed used by the subject and a water sensor capable of detecting the water intake of the subject. The second sensor is the data processing device according to Configuration 6 or 7, including at least one of an excretion device used by the subject and an excretion sensor capable of detecting the subject's excretion.

[0114] (Configuration 9) The first event includes at least one of eating and drinking, body movement, and change in posture. The predicted state of the subject at the time of occurrence of the first event is the data processing device according to Configuration 1, including prediction of aspiration.

[0115] (Configuration 10) The state prediction information is the data processing device according to Configuration 9, including prediction of the first event different from normal.

[0116] (Configuration 11) The first sensor includes at least one of a body movement sensor provided on the bed used by the subject, a food and drink sensor capable of detecting the subject's food and drink, and a posture sensor capable of detecting the subject's posture. The second sensor is the data processing device according to Configuration 9 or 10, including a swallowing sensor capable of detecting abnormal swallowing of the subject.

[0117] (Configuration 12) The first data information A state where the signal intensity corresponding to the body movement of the subject exceeds a first threshold value The frequency of the state where the signal intensity exceeds the first threshold value The occurrence time of the state where the signal intensity exceeds the first threshold value, and The duration of the state where the signal intensity exceeds the first threshold value is the data processing device according to any one of Configurations 2 to 11, including at least one of the above.

[0118] (Configuration 13) The processing unit can further perform a second operation. In the second operation, the processing unit can acquire first information obtained from the first sensor and second information obtained from the second sensor. The amount of the first information is larger than the amount of the second information. The second information includes learning event occurrence information regarding the occurrence of a learning event detected by the second sensor, and the learning event corresponds to the first event. The first information includes first period information obtained from the first sensor during a first period that includes a time before the occurrence of the learning event. In the second operation, the processing unit can derive a machine learning model using the learning event occurrence information and information based on the first period information as teacher data. The data processing device according to any one of Configurations 1 to 12, wherein the processing unit can perform the first operation based on the machine learning model.

[0119] (Configuration 14) In the second operation, the processing unit adds a first annotation to the first period information based on the learning event occurrence information. The data processing device according to Configuration 13, wherein the teacher data includes the first period information to which the first annotation is added.

[0120] (Configuration 15) A part of the first data is obtained from a third sensor capable of detecting the state of the subject. The first sensor includes a body movement sensor provided on a bed used by the subject. The third sensor includes an imaging device that images the subject. The second sensor includes at least one of an excretion device used by the subject and an excretion sensor capable of detecting the excretion of the subject. The first event includes excretion. The data processing device according to Configuration 1, wherein the state at the time of the occurrence of the first event includes at least a prediction of either a fall or a tumble.

[0121] (Configuration 16) A part of the first data is obtained from a third sensor capable of detecting the state of the subject, The first sensor includes a body movement sensor provided on a bed used by the subject, The third sensor includes an imaging device for imaging the subject, The second sensor includes a weight sensor capable of detecting the weight from the subject, The first event includes getting out of bed, The state at the time of the occurrence of the first event includes at least a prediction of either a fall or a tumble, the data processing device according to Configuration 1.

[0122] (Configuration 17) The processing unit can further perform a second operation, In the second operation, the processing unit can acquire first information obtained from the first sensor, second information obtained from the second sensor, and third information obtained from the third sensor, The amount of the first information is larger than the amount of the second information, The second information includes second event occurrence information regarding the occurrence of a second event detected by the second sensor, and the second event corresponds to the first event, The first information includes first period information obtained from the first sensor during a first period including before the occurrence of the second event, The third information includes second period information obtained from the third sensor during a second period including before the occurrence of the second event, In the second operation, the processing unit can derive a machine learning model using the second event occurrence information, information based on the first period information, and information based on the second period information as teacher data, The processing unit can perform the first operation based on the machine learning model, the data processing device according to Configuration 15 or 16.

[0123] (Configuration 18) In the second operation, the processing unit adds a first annotation to the first period information and a second annotation to the second period information based on the second event occurrence information, The data processing apparatus according to Configuration 17, wherein the teacher data includes the first period information to which the first annotation is added and the second period information to which the second annotation is added.

[0124] (Configuration 19) an acquisition unit capable of acquiring first data; a processing unit capable of performing a first operation based on the first data acquired by the acquisition unit; and at least a part of the first data is obtained from a first sensor capable of detecting a state of a target person, In the first operation, when the first data includes first data information, the processing unit can output first event prediction information regarding the occurrence of a first event regarding the target person and second event prediction information regarding the occurrence of a second event after the first event, The occurrence of the first event can be detected by a second sensor capable of detecting the state of the target person, The data processing apparatus, wherein the occurrence of the second event can be detected by a third sensor capable of detecting the state of the target person.

[0125] (Configuration 20) The first event includes getting out of bed, The data processing apparatus according to Configuration 19, wherein the second event includes excretion.

[0126] (Configuration 21) The first sensor includes a body movement sensor provided on a bed used by the target person, The second sensor includes a weight sensor capable of detecting a weight from the target person, The data processing apparatus according to Configuration 19 or 20, wherein the third sensor includes at least one of an excretion device used by the target person and an excretion sensor capable of detecting excretion of the target person.

[0127] (Configuration 22) The data processing device according to Configuration 1, the first sensor, a terminal device, and is provided with, the terminal device includes an output unit, the output unit is a care support system capable of outputting the first event prediction information and the state prediction information.

[0128] (Configuration 23) Cause the processing unit to obtain first data, Cause the processing unit to perform a first operation based on the first data, At least a part of the first data is obtained from a first sensor capable of detecting the state of the subject, In the first operation, when the first data includes first data information, the processing unit can output first event prediction information regarding the occurrence of a first event regarding the subject and state prediction information regarding the predicted state of the subject at the time of the occurrence of the first event, The occurrence of the first event can be detected by a second sensor capable of detecting the state of the subject, The predicted state of the subject at the time of the occurrence of the first event is not detected by the second sensor, a data processing method.

[0129] According to the embodiment, a data processing device, a care support system, and a data processing method capable of more appropriate processing can be provided.

[0130] As described above, the embodiments of the present invention have been described with reference to specific examples. However, the present invention is not limited to these specific examples. For example, regarding the specific configuration of each element such as the processing unit and the acquisition unit included in the data processing device, the present invention can be similarly implemented by appropriately selecting from the range known to those skilled in the art, and as long as the same effects can be obtained, it is included in the scope of the present invention.

[0131] In addition, combinations of any two or more elements of each specific example, as long as they are technically possible, are also included in the scope of the present invention as long as they fall within the gist of the present invention.

[0132] In addition, based on the data processing device, care support system, and data processing method described above as embodiments of the present invention, all data processing devices, care support systems, and data processing methods that can be appropriately designed and modified by those skilled in the art also belong to the scope of the present invention as long as they fall within the gist of the present invention.

[0133] In addition, within the scope of the idea of the present invention, those skilled in the art can conceive of various modification examples and correction examples, and it is understood that those modification examples and correction examples also belong to the scope of the present invention.

Explanation of Reference Numerals

[0134] 31 to 33... First to third sensors, 41... Body movement sensor, 42... Imaging device, 43... Load sensor, 44... Excretion device, 45... Excretion sensor, 46... Moisture sensor, 47... Eating and drinking sensor, 47a... Image sensor, 48... Posture sensor, 49... Swallowing sensor, 70... Data processing device, 71... Processing unit, 72... Acquisition unit, 73... Storage unit, 75... Communication unit, 76... Storage unit, 79b... Display unit, 79c... Input unit, 81... Terminal device, 82... Output unit, 85... Subject, 86... Bed, 88... Machine learning model, 110... Care support system, DA1... First data, DI1... First data information, I1 to I3... First to third information, IL0... Learning event occurrence information, IL2... Second event occurrence information, IP1... First event prediction information, IS1... Signal strength, IS2... Strength, IS3... Signal strength, Itp1, Itp2... First and second period information, O01... Output information, OP1, OP2... First and second operations, SP1... State prediction information, SS1... Signal strength, SS2... Strength, SS3... Signal strength, SV1... First threshold, tL0... Time, tL2... Time, tc1... Frequency, ti1... State, tm... Time, tm1... Occurrence time, tm2... Time, td1... Duration, tp1, tp2... First and second periods

Claims

1. An acquisition unit capable of acquiring first data; A processing unit capable of performing a first operation based on the first data acquired by the acquisition unit; Comprising: At least a part of the first data is obtained from a first sensor capable of detecting the state of the target person; In the first operation, when the first data includes first data information, the processing unit can output first event prediction information regarding the occurrence of a first event regarding the target person and state prediction information regarding the predicted state of the target person at the time of the occurrence of the first event; The occurrence of the first event can be detected by a second sensor capable of detecting the state of the target person; The predicted state of the target person at the time of the occurrence of the first event is not detected by the second sensor; Based on the second data acquired from the second sensor, annotation of the first event prediction information and the state prediction information is added to the first data acquired from the first sensor, and a learning model is generated based on the first data and the annotation, a data processing device.

2. The first event includes at least one of getting out of bed and excretion; The predicted state of the target person at the time of the occurrence of the first event includes at least one of prediction of falling and tumbling, the data processing device according to claim 1.

3. The first event includes at least one of water intake, body movement and excretion; The predicted state of the target person at the time of the occurrence of the first event includes prediction of incontinence, the data processing device according to claim 1.

4. The processing unit can further perform a second operation; In the second operation, the processing unit can acquire first information obtained from the first sensor and second information obtained from the second sensor; The amount of the first information is larger than the amount of the second information; The second information includes learning-time event occurrence information regarding the occurrence of a learning-time event detected by the second sensor, and the learning-time event corresponds to the first event; The first information includes first period information obtained from the first sensor during a first period including before the occurrence of the learning-time event; In the second operation, the processing unit can derive the learning model using the learning-time event occurrence information and the information based on the first period information as teacher data. The data processing apparatus according to claim 1, wherein the processing unit is capable of performing the first operation based on the learning model.

5. In the second operation, the processing unit adds a first annotation to the first period information based on the learning event occurrence information, The data processing apparatus according to claim 4, wherein the teacher data includes the first period information to which the first annotation is added.

6. A part of the first data is obtained from a third sensor capable of detecting the state of the subject, The first sensor includes a body movement sensor provided on a bed used by the subject, The third sensor includes an imaging device that images the subject, The second sensor includes at least one of an excretion device used by the subject and an excretion sensor capable of detecting the excretion of the subject, The first event includes excretion, The data processing apparatus according to claim 1, wherein the state at the time of the occurrence of the first event includes at least a prediction of either a fall or a tumble.

7. The processing unit is further capable of performing a second operation, In the second operation, the processing unit can acquire first information obtained from the first sensor, second information obtained from the second sensor, and third information obtained from the third sensor, The amount of the first information is larger than the amount of the second information, The second information includes second event occurrence information regarding the occurrence of a second event detected by the second sensor, and the second event corresponds to the first event, The first information includes first period information obtained from the first sensor during a first period including before the occurrence of the second event, The third information includes second period information obtained from the third sensor during a second period including before the occurrence of the second event, In the second operation, the processing unit can derive the learning model using the second event occurrence information, information based on the first period information, and information based on the second period information as teacher data, The data processing apparatus according to claim 6, wherein the processing unit is capable of performing the first operation based on the learning model.

8. In the second operation, the processing unit adds a first annotation to the first period information and a second annotation to the second period information based on the second event occurrence information, The teacher data includes the first period information with the first annotation added and the second period information with the second annotation added, the data processing apparatus according to claim 7.

9. A data processing apparatus according to any one of claims 1 to 8, the first sensor, a terminal device, and comprising: the terminal device includes an output unit, the output unit is capable of outputting the first event prediction information and the state prediction information, a care support system.

10. The processing unit obtains first data, the processing unit performs a first operation based on the first data, at least a part of the first data is obtained from a first sensor capable of detecting the state of a target person, in the first operation, when the first data includes first data information, the processing unit can output first event prediction information regarding the occurrence of a first event regarding the target person and state prediction information regarding the predicted state of the target person at the time of the occurrence of the first event, the occurrence of the first event can be detected by a second sensor capable of detecting the state of the target person, the predicted state of the target person at the time of the occurrence of the first event is not detected by the second sensor, Based on the second data obtained from the second sensor, the first data obtained from the first sensor is annotated with the first event prediction information and the state prediction information, and a learning model is generated based on the first data and the annotation, a data processing method.

Images