Information processing apparatus, information processing method, and program

Abstract

To present an action that is effective for alleviating the symptom of a lifestyle-related disease and maintaining and improving body conditions, and that can be easily executed by a subject.SOLUTION: An information processing apparatus according to an embodiment comprises an acquisition unit, a setting unit, and a presentation unit. A first acquisition unit acquires current state information representing the current state of a subject. The setting unit sets a target state representing a state targeted by the subject. The presentation unit presents recommended action information representing an action recommended for the subject on the basis of the current state information and the target state.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The embodiments disclosed in the present specification and drawings relate to an information processing device, an information processing method, and a program. [Background technology]

[0002] In recent years, the number of patients with lifestyle-related diseases, including diabetes, has been increasing, becoming a social problem. The progression of symptoms of such lifestyle-related diseases is greatly influenced by the patient's daily lifestyle. Conventionally, there are known techniques for setting behavioral goals for patients to prevent and improve lifestyle-related diseases based on the patient's personality, lifestyle, etc.

[0003] However, due to the diversity of patients' values, lifestyles, genetic diversity, etc., there is a possibility that patients may not take the actions necessary to achieve the set goals, or that even if they do take the actions, they may not be very effective. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2022-117339 Summary of the Invention [Problem to be solved by the invention]

[0005] One of the problems to be solved by the embodiments disclosed in this specification and drawings is to present an action that is effective in alleviating the symptoms of lifestyle-related diseases and maintaining and improving physical condition and that is easy for the subject to perform. However, the problem to be solved by the embodiments disclosed in this specification and drawings is not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described below can also be positioned as other problems. [Means for solving the problem]

[0006] An information processing device according to an embodiment includes an acquisition unit, a setting unit, and a presentation unit. The acquisition unit acquires current state information representing a current state of a subject. The setting unit sets a target state representing a target state of the subject. The presentation unit presents recommended behavior information representing a behavior recommended for the subject based on the current state information and the target state. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a medical support system according to the first embodiment. [Figure 2] FIG. 2 is a block diagram showing an example of the configuration of the medical support device according to the first embodiment. [Figure 3] FIG. 3 is a diagram illustrating an example of a calculation result of the feasible level according to the first embodiment. [Figure 4] FIG. 4 is a diagram illustrating an example of an estimation result of an improvement effect according to the first embodiment. [Figure 5] FIG. 5 is a diagram illustrating an example of a result of identifying a recommended action according to the first embodiment. [Figure 6] FIG. 6 is a flowchart showing an example of processing executed by the medical support device according to the first embodiment. [Figure 7] FIG. 7 is a block diagram showing an example of the configuration of a medical support device according to the second embodiment. [Figure 8] FIG. 8 is a diagram illustrating an example of a data configuration of correspondence information according to the second embodiment. [Figure 9] FIG. 9 is a diagram illustrating an example of a result of identifying a behavioral barrier according to the second embodiment. [Figure 10] FIG. 10 is a flowchart showing an example of processing executed by the medical support device according to the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, embodiments of an information processing device, an information processing method, and a program will be described in detail with reference to the drawings.

[0009] (First embodiment) Fig. 1 is a diagram showing an example of the configuration of a medical support system S according to the first embodiment. As shown in Fig. 1, the medical support system S includes a medical support device 100, a medical information storage device 200, a medical interview system 201, and an electronic medical record system 202. The medical support device 100 is an example of an information processing device. These components are communicably connected via a network N such as an in-hospital LAN (Local Area Network) or the Internet.

[0010] The medical support device 100 of this embodiment is an information processing device that supports medical treatment for a subject to be managed, which represents a patient with a lifestyle-related disease or a person at risk of developing a lifestyle-related disease. The subject to be managed is an example of a test subject. The medical support device 100 is realized by, for example, a server device or a PC (Personal Computer). The functions of the medical support device 100 will be described in detail later.

[0011] The medical information storage device 200 stores, for example, a PHR (Personal Health Record) 20a. The PHR 20a stores, for example, information related to an individual's medical care. The PHR 20a may include information collected by a health monitoring device that monitors the heart rate, electrocardiogram, blood pressure, etc. of the person to be managed. In this case, the information stored as the PHR 20a is an example of monitor information.

[0012] The heart rate, electrocardiogram, blood pressure, etc. of the managed person represent the physical activity of the managed person and are examples of activity information. Hereinafter, data on the heart rate, electrocardiogram, blood pressure, etc. of the managed person will also be referred to as information on the physical activity of the managed person. In this case, the medical information storage device 200 may receive data from a health monitoring device, etc., communicably connected via the network N, and store the data in the PHR 20a.

[0013] The medical interview system 201 is a device that stores medical interview information indicating the contents of the responses of the person to a medical interview. For example, the medical interview information is information including responses to the current condition (including symptoms) of the person to be managed, responses to the living environment of the person to be managed, and responses to a personality diagnosis of the person to be managed.

[0014] The electronic medical record system 202 is a device that stores electronic medical records.

[0015] In terms of hardware, the medical information storage device 200, the medical interview system 201, and the electronic medical record system 202 are, for example, computers such as server devices. Note that some of these devices may be integrated. For example, the functions of the multiple devices shown in Fig. 1 may be integrated as an HIS (Hospital Information System).

[0016] The user terminal 203 is a device that inputs information to the medical support device 100 and displays information output from the medical support device 100. The user terminal 203 is an example of a display device. The user terminal 203 is realized by, for example, a smartphone, a tablet terminal, or a laptop PC.

[0017] Note that the configuration shown in Fig. 1 is an example, and the medical support system S may further include other components, or may not include some of the components shown in Fig. 1. For example, the medical support system S may include a pharmacy department system, etc.

[0018] Next, a description will be given of the functions of the medical support device 100. Fig. 2 is a block diagram showing an example of the configuration of the medical support device 100 according to the first embodiment.

[0019] The medical support device 100 includes a network interface 110, a memory 120, an input interface 130, a display 140, and a processing circuit 150.

[0020] The NW interface 110 is connected to the processing circuit 150 and controls the transmission and communication of various data between the medical support device 100 and other information processing devices. The NW interface 110 is realized by a network card, a network adapter, a NIC (Network Interface Controller), or the like.

[0021] The memory 120 stores in advance various types of information used by the processing circuit 150. For example, the memory 120 of this embodiment stores a feasible level calculation model 121, an effect estimation model 122, and a behavior list 123.

[0022] The feasibility level calculation model 121 is a generation AI that generates sentences related to feasibility levels that indicate the range of actions that a managed user can perform. The feasibility level calculation model 121 is an example of a first trained model. For example, the feasibility level calculation model 121 is a large language model (LLM). Here, LLM refers to a language model constructed using a large amount of data and well-known deep learning technology.

[0023] As an example, the feasibility level calculation model 121 generates a sentence about the feasibility level in response to input of current status information indicating the current status of the managed person and the action. The current status information will be described later.

[0024] The effect prediction model 122 is a generation AI that generates sentences regarding improvement effects that indicate the degree to which the condition of the managed person will improve by performing an action. The effect prediction model 122 is an example of a second trained model. For example, the effect prediction model 122 is an LLM. As an example, the effect prediction model 122 generates sentences regarding improvement effects expected by performing an action in response to input of current state information and an action. Improvement effects will be described later.

[0025] The behavior list 123 is a list in which a plurality of behaviors are registered. For example, the behavior list 123 is information representing a plurality of behaviors by the type of behavior and the strength of the type of behavior.

[0026] For example, types of behavior include walking, jogging, diet (e.g., restricting calorie intake), cycling, swimming, etc. Furthermore, for example, the intensity of walking can be "one day a week, up to 5 minutes," "three days a week, 6 to 30 minutes," "five days or more a week, 31 minutes or more," etc. One behavior is expressed as "walking: one day a week, up to 5 minutes."

[0027] The memory 120 also stores various programs in addition to the above. The memory 120 is realized by, for example, a random access memory (RAM), a semiconductor memory element such as a flash memory, an optical disk, a hard disk drive (HDD), or a solid state drive (SSD). The memory 120 may be located within the medical support device 100, or may be located in an external storage device connected to the medical support device 100 via a network N, for example.

[0028] The input interface 130 is realized by a trackball, switch buttons, a mouse, a keyboard, a touchpad that performs input operations by touching the operation surface, a touchscreen that integrates a display screen and a touchpad, a non-contact input circuit using an optical sensor, and a voice input circuit, etc.

[0029] The input interface 130 is connected to the processing circuitry 150, converts input operations received from an operator into electrical signals, and outputs the signals to the processing circuitry 150. In this specification, the input interface 130 is not limited to an interface equipped with physical operation components such as a mouse and a keyboard. For example, an example of the input interface 130 also includes an electrical signal processing circuit that receives electrical signals corresponding to input operations from an external input device provided separately from the medical support device 100 and outputs the electrical signals to the processing circuitry 150.

[0030] The display 140 is a dot-matrix electronic display such as a liquid crystal display or an organic electro-luminescence (OEL) display. The input interface 130 and the display 140 may be integrated.

[0031] For example, the input interface 130 and the display 140 may be realized by a touch panel. The display 140 is an example of a display unit in this embodiment. The display 140 may be provided separately from the medical support device 100 and connected to the medical support device 100 via a network N, for example.

[0032] The processing circuitry 150 is a processor that realizes functions corresponding to each program by reading and executing the programs from the memory 120. The processing circuitry 150 of this embodiment includes a learning function 151, an acquisition function 152, a setting function 153, a calculation function 154, an estimation function 155, a range identification function 156, a behavior identification function 157, a presentation function 158, and a display control function 159.

[0033] The acquisition function 152 is an example of an acquisition unit. The setting function 153 is an example of a setting unit. The calculation function 154 is an example of a calculation unit. The range identification function 156 is an example of a range identification unit. The behavior identification function 157 is an example of a behavior identification unit. The presentation function 158 and the display control function 159 are examples of a presentation unit.

[0034] Here, for example, each of the processing functions of the processing circuit 150, namely, a learning function 151, an acquisition function 152, a setting function 153, a calculation function 154, an estimation function 155, a range identification function 156, a behavior identification function 157, a presentation function 158, and a display control function 159, is stored in the memory 120 in the form of a program executable by a computer. The processing circuit 150 is a processor.

[0035] For example, the processing circuitry 150 realizes the functions corresponding to each program by reading and executing the programs from the memory 120. In other words, after reading each program, the processing circuitry 150 has the functions shown in the processing circuitry 150 of FIG.

[0036] In FIG. 2, it has been explained that the processing functions performed by the learning function 151, the acquisition function 152, the setting function 153, the calculation function 154, the estimation function 155, the range identification function 156, the behavior identification function 157, the presentation function 158, and the display control function 159 are realized by a single processor, but it is also possible to combine multiple independent processors to form the processing circuit 150, and realize the functions by each processor executing a program.

[0037] Also, in Figure 2, a single memory 120 is described as storing programs corresponding to each processing function, but multiple memories may be distributed and arranged, and the processing circuit 150 may read corresponding programs from individual memories.

[0038] In the above description, an example has been described in which a "processor" reads out a program corresponding to each function from a memory and executes it, but the embodiment is not limited to this. The term "processor" refers to a circuit such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an Application Specific Integrated Circuit (ASIC), a programmable logic device (e.g., a Simple Programmable Logic Device (SPLD), a Complex Programmable Logic Device (CPLD), and a Field Programmable Gate Array (FPGA)).

[0039] If the processor is a CPU, for example, the processor realizes its functions by reading and executing a program stored in memory 120. On the other hand, if the processor is an ASIC, instead of storing a program in memory 120, the function is directly incorporated as a logic circuit within the circuitry of the processor.

[0040] Note that each processor in this embodiment is not limited to being configured as a single circuit, but may be configured as a single processor by combining multiple independent circuits to realize its functions. Furthermore, multiple components in Figure 2 may be integrated into a single processor to realize its functions.

[0041] The medical support device 100 of this embodiment has a function of presenting to the user recommended actions for improving the condition (symptoms) of the person to be managed. Details of each function of the medical support device 100 will be described below.

[0042] The learning function 151 learns various models using deep learning technology. For example, the learning function 151 learns the feasible level calculation model 121 and the effect estimation model 122. First, the learning of the feasible level calculation model 121 will be described.

[0043] As an example, the learning function 151 collects text describing the current status information of a large number of managed individuals. Here, the current status information is information that represents the status of the managed individuals related to lifestyle-related diseases. For example, the current status information is information that includes lifestyle information, vital information, and attribute information of the managed individuals.

[0044] The lifestyle information is information about the lifestyle of the person to be managed. For example, the lifestyle information is information about the exercise and diet that the person to be managed engages in in their daily lives. As an example, the learning function 151 acquires medical interview information from the medical interview system 201 via the NW interface 110. The learning function 151 extracts information about exercise and diet from the medical interview information to collect text describing the lifestyle information of the person to be managed.

[0045] Vital information is information relating to indicators showing the condition of the managed person. For example, vital information is information such as the pulse (heart rate), respiration, blood pressure, and body temperature of the managed person, as well as information showing the test results of various tests on the managed person. As an example, the learning function 151 acquires information such as the pulse (heart rate), respiration, blood pressure, and body temperature of the managed person from the PHR 20a of the medical information storage device 200 via the NW interface 110.

[0046] Similarly to the above, the learning function 151 acquires the electronic medical record of the managed person from the electronic medical record system 202. The learning function 151 extracts information such as the pulse (heart rate), breathing, blood pressure, and body temperature of the managed person, as well as information indicating the test results of various tests on the managed person, from the acquired information, thereby collecting text describing the vital information of the managed person.

[0047] The attribute information is information relating to the attributes of the managed person. For example, the attribute information is information such as the gender, age, height, and weight of the managed person. As an example, the learning function 151 collects text describing the attribute information of the managed person by extracting information such as the gender, age, height, and weight of the managed person from the acquired electronic medical record.

[0048] The learning function 151 generates a set of input training data, including text describing the collected lifestyle information of the managed person, text describing vital information of the managed person, text describing current status information including text describing attribute information of the managed person, and text describing actions registered in the action list 123.

[0049] Furthermore, the learning function 151 generates text data describing the degree of difficulty of performing a behavior registered in the behavior list 123, the degree of difficulty indicating the degree of difficulty in performing the behavior corresponding to the current state information of the managed user.

[0050] For example, the learning function 151 generates text data (output teacher data) describing the difficulty level by correcting the standard difficulty level that is set in advance for each action according to the attribute information of the managed person with the living information and vital information of the managed person.

[0051] As an example, if the behavior is "walking: 3 days a week for 5 to 15 minutes" and the standard difficulty level is 2, and the person to be managed normally walks for 5 minutes or less once a week, the learning function 151 will correct the difficulty level by -1. Also, if the vital information of the person to be managed is in a worse state (for example, high blood pressure) than someone with the same attributes, the learning function 151 will correct the difficulty level by +1.

[0052] The learning function 151 may output the training data as text (numerical values) input by the user as the difficulty level corresponding to the input training data.

[0053] The learning function 151 also generates a set of the generated input training data and the generated output training data as a training data set. The learning function 151 uses the generated training data set to train the feasibility level calculation model 121 using a known deep learning technique.

[0054] Next, we will explain the learning of the effect estimation model 122. For example, the learning function 151 collects text describing the improvement effect corresponding to the current state information of the managed person by performing the action registered in the action list 123. For example, the improvement effect is information that indicates the degree of improvement in the state (symptoms) of the managed person.

[0055] As an example, the learning function 151 acquires an electronic medical record from the electronic medical record system 202 via the NW interface 110. The learning function 151 extracts, from the acquired electronic medical record, a text describing the performed action, a text describing current state information before the action was performed, and a text describing the improvement effect due to the action being performed.

[0056] The learning function 151 generates, as input training data, a set of text describing the performed action and text describing current state information before the action was performed from the extracted text. Furthermore, the learning function 151 generates, as output training data, text describing the improvement effect quantified based on the text describing the improvement effect due to the action execution.

[0057] As an example, the learning function 151 uses a known natural language processing technique to analyze the extracted text describing the improvement effect due to the execution of the behavior, and quantifies the degree of improvement in the condition of the managed person due to the execution of the behavior. More specifically, the learning function 151 may quantify the degree of improvement in the condition of the managed person based on the degree of change in the test results (test values) of various tests included in the managed person's regular health checkup during the period when the behavior is being executed.

[0058] The learning function 151 may output the training data as text (numerical values) input by the user as an improvement effect corresponding to the input training data.

[0059] The learning function 151 also generates a set of the generated input teacher data and the generated output teacher data as a learning dataset. The learning function 151 uses the generated learning dataset to train the effect prediction model 122 by a known deep learning technique.

[0060] The acquisition function 152 acquires the current status information of the managed entity. In this embodiment, the current status information is expressed in text.

[0061] For example, the acquisition function 152 acquires medical interview information of the managed person from the medical interview system 201 via the NW interface 110. The acquisition function 152 acquires lifestyle information of the managed person included in the current condition information of the managed person by extracting information on exercise and information on diet from the medical interview information of the managed person.

[0062] Also, for example, the acquisition function 152 acquires information such as the pulse (heart rate), breathing, blood pressure, body temperature, etc. of the person to be managed from the PHR 20a of the medical information storage device 200 via the NW interface 110, and acquires the electronic medical record of the person to be managed from the electronic medical record system 202.

[0063] The acquisition function 152 acquires vital information of the managed person contained in the current status information of the managed person by extracting information such as the managed person's pulse (heart rate), breathing, blood pressure, body temperature, etc., and information representing the test results of various tests on the managed person from the above information.

[0064] Also, for example, the acquisition function 152 acquires attribute information of the managed person included in the current status information of the managed person by extracting information such as the gender, age, height, and weight of the managed person from the managed person's electronic medical record.

[0065] The setting function 153 sets a target state that represents a state that the person to be managed is aiming for. For example, in the case of managing diabetes, the target state indicates a state that the person to be managed is aiming for, from among states classified as normal, borderline, mild, moderate, severe, etc.

[0066] Here, as an example, the setting function 153 performs scoring based on test values ​​such as fasting blood glucose level and hemoglobin A1c, symptoms (subjective symptoms, objective symptoms), etc., and classifies the condition of the person to be managed into normal range, borderline range, mild condition, moderate condition, severe condition, etc. depending on the scoring results.

[0067] The above is just an example, and any method for classifying conditions, scoring methods, etc. may be set. For example, conditions may be classified using various medical guidelines related to lifestyle-related diseases.

[0068] For example, the setting function 153 refers to the electronic medical record of the managed person via the NW interface 110, extracts words related to the target state of the managed person, and sets the target state based on the extraction result. Note that the setting function 153 may set the target state according to user input.

[0069] The calculation function 154 calculates an executable level for each of a plurality of actions based on the current state information acquired by the acquisition function 152. The executable level is an example of the degree of difficulty of execution.

[0070] For example, the calculation function 154 inputs the current state information (text) acquired by the acquisition function 152 and text describing the names of the behaviors registered in the behavior list 123 to the feasibility level calculation model 121. The calculation function 154 calculates the feasibility level of the managed user for each behavior registered in the behavior list 123 according to the output of the feasibility level calculation model 121.

[0071] 3 is a diagram illustrating an example of a calculation result of the feasible level according to the first embodiment. FIG. 3 shows an example of a calculation result of the feasible levels of walking, jogging, eating (calorie intake restriction), cycling, and swimming for a specific managed person, calculated by the calculation function 154.

[0072] The horizontal axis in Figure 3 represents the feasibility level. The more difficult the behavior is for the person being managed, the more difficult it is (the more difficult it is), and the more difficult it is for the person being managed, the more difficult it is (the more difficult it is).

[0073] The explanation will be continued by returning to Fig. 2. The estimation function 155 estimates the improvement effect obtained by performing an action based on the current state information acquired by the acquisition function 152.

[0074] For example, the estimation function 155 inputs the current state information acquired by the acquisition function 152 and text describing the names of actions registered in the action list 123 into the effect estimation model 122. The estimation function 155 estimates the improvement effect when each action registered in the action list 123 is performed according to the output of the effect estimation model 122.

[0075] 4 is a diagram illustrating an example of the estimation result of the effect according to the first embodiment. Fig. 4 shows an example of the estimation result of the improvement effect when a specific managed person performs walking, jogging, eating, cycling, and swimming, by the estimation function 155.

[0076] The vertical axis in Figure 4 represents the improvement effect. The improvement effect in Figure 4 shows how close it is estimated that the managed person's condition will be brought to normal by taking action, with the origin being the current condition of the managed person and the highest point being the normal condition. The closer to the "high" side in the figure, the greater the effect of improving the managed person's condition, and the closer to the "low" side in the figure, the less effective it will be in improving the managed person's condition.

[0077] Returning to Fig. 2, the explanation will be continued. The range identification function 156 identifies the execution range of an action based on the current state information acquired by the acquisition function 152. The execution range is the range of actions that the managed person is currently performing. In this embodiment, the execution range includes the range of actions that are not actually performed but that are presumed to be executable by the managed person.

[0078] For example, the range identification function 156 extracts words related to the execution of the type of behavior registered in the behavior list 123 from the current state information acquired by the acquisition function 152. The range identification function 156 identifies the execution range for the type of behavior registered in the behavior list 123 according to the extraction result.

[0079] Furthermore, the range identification function 156 identifies the range of execution for the type of behavior for which words related to execution could not be extracted from the current state information acquired by the acquisition function 152, in accordance with the affinity between the current state information of the managed person and the type of behavior for which words related to execution could not be extracted from the current state information.

[0080] Specifically, when the type of activity is cycling, the range identification function 156 calculates the affinity by scoring the affinity based on conditions such as owning a bicycle, having little elevation change around the address, and having a workplace within XX kilometers of the home.The range identification function 156 identifies the range of the activity corresponding to the affinity calculation result as the execution range based on the correspondence information associating the affinity with the range of the activity and the affinity calculation result.

[0081] In addition, when the type of activity is swimming, the range identification function 156 performs scoring based on conditions such as having a pool in the living area, owning a swimsuit, and having learned to swim, and identifies the execution range in the same manner as above.

[0082] In this embodiment, the range specification function 156 specifies the execution range of an action based on the average value of the maximum feasible level in the execution range of each action type. Note that the range specification function 156 may specify the execution range for each action type, may specify the execution range according to the action type with the narrowest execution range, or may specify the execution range according to the action type with the widest execution range.

[0083] The behavior specifying function 157 specifies a recommended behavior that indicates a behavior that is recommended to the managed user based on the current state information acquired by the acquisition function 152 and the target state set by the setting function 153 .

[0084] For example, the behavior identification function 157 first identifies, from among the multiple behaviors registered in the behavior list 123, a behavior that is estimated to be capable of reaching the goal state, based on the goal state set by the setting function 153 and the improvement effect estimated by the estimation function 155. The behavior identified in this case is an example of a behavior candidate.

[0085] Next, the behavior specifying function 157 specifies, as a recommended behavior, a behavior whose executable level is close to the executable range specified by the range specifying function 156, from among the behaviors that are estimated to be capable of reaching the specified goal state.

[0086] Specifically, the behavior identification function 157 selects three behaviors from among the behaviors that are estimated to be capable of reaching the goal state in order of feasibility level that is closest to the feasibility range, and identifies the selected behaviors as recommended behaviors. Note that, although three behaviors are selected in order of feasibility level that is closest to the feasibility range in the above, the number of selected behaviors is not limited to three. For example, the number of selected behaviors may be two or less, or four or more.

[0087] FIG. 5 is a diagram illustrating an example of a result of identifying a recommended action according to the first embodiment.

[0088] 5, the range identification function 156 identifies E, which is the average value of the maximum feasible levels in the execution range of each type of activity (walking, jogging, eating, cycling, swimming) on ​​the graph, from the current state information acquired by the acquisition function 152. Then, the range identification function 156 identifies, on the horizontal axis of the graph in FIG. 5, the range of activities whose feasible levels are equal to or less than E, as the execution range of the activities.

[0089] 5, the behavior identification function 157 identifies T, which is a value of an improvement effect that can achieve the target state on the graph, from the target state set by the setting function 153. Then, the behavior identification function 157 identifies, on the vertical axis of the graph in FIG. 5, a range of behaviors whose improvement effect is equal to or greater than T, as a range of behaviors that can reach the target state.

[0090] Furthermore, the behavior identification function 157 selects three behaviors from among the behaviors that have an improvement effect of T or more, in order of the feasibility level closest to E. In the example of FIG. 5, the behavior identification function 157 identifies the selected "swimming: 3 days a week, 16 to 30 minutes," "walking: 5 days or more a week, 31 minutes or more," and "jogging: 3 days a week, 6 to 15 minutes" as recommended behaviors.

[0091] Returning to Fig. 2, the explanation will be continued. The presentation function 158 presents recommended action information related to the recommended action identified by the action identification function 157. For example, the presentation function 158 generates a presentation screen for presenting the graph shown in Fig. 5 to the user as recommended action information.

[0092] The display control function 159 displays various types of information on a display device. For example, the display control function 159 controls the display of the user terminal 203 to display the presentation screen generated by the presentation function 158 via the NW interface 110. The display control function 159 may also display the generated presentation screen on the display 140.

[0093] Next, a description will be given of the processing executed by the medical support device 100. Fig. 6 is a flowchart showing an example of the processing executed by the medical support device 100 according to the first embodiment.

[0094] First, the acquisition function 152 acquires current status information of the managed person (step ST101). For example, the acquisition function 152 acquires the medical interview information and electronic medical record of the managed person from the medical interview system 201 and the electronic medical record system 202 via the NW interface 110. The acquisition function 152 acquires information including lifestyle information, vital information, and attribute information extracted from the medical interview information and electronic medical record of the managed person as current status information.

[0095] Next, the setting function 153 sets a target state of the managed person (step ST102). For example, the setting function 153 refers to the electronic medical record of the managed person stored in the electronic medical record system 202 via the NW interface 110, and extracts words related to the target state. The setting function 153 sets the target state based on the extracted words.

[0096] Next, the calculation function 154 calculates the feasibility level of the behavior of the managed individual (step ST103).

[0097] For example, the calculation function 154 inputs the current state information acquired in step ST101 and text describing the names of the actions registered in the action list 123 to the feasible level calculation model 121 stored in the memory 120. The calculation function 154 calculates the feasible level of each action registered in the action list 123 according to the output result of the feasible level calculation model 121.

[0098] Next, the estimation function 155 estimates the improvement effect due to the execution of the behavior of the managed person (step ST104).

[0099] For example, the estimation function 155 inputs the current state information acquired in step ST101 and text describing the names of the actions registered in the action list 123 into the effect estimation model 122 stored in the memory 120. The estimation function 155 estimates the improvement effect resulting from the execution of each action registered in the action list 123 according to the output result of the effect estimation model 122.

[0100] Next, the range identification function 156 identifies the execution range of the behavior of the managed user (step ST105). For example, the range identification function 156 extracts words related to the execution of the behavior type registered in the behavior list 123 from the current state information acquired in step ST101, and identifies the execution range of the behavior on the graph shown in FIG. 4 from the extraction result.

[0101] Furthermore, for example, for a type of behavior for which no words related to execution could be extracted from the current state information acquired in step ST101, the range identification function 156 identifies the execution range on the graph shown in Figure 4 based on the affinity between the current state information of the managed person and the type of behavior for which no words related to execution could be extracted from the current state information.

[0102] Next, the behavior identification function 157 identifies a range of behaviors that can reach the goal state (step ST106). For example, the behavior identification function 157 identifies a range of behaviors that can reach the goal state on the graph shown in FIG. 4, based on the goal state set in step ST102 and the improvement effect estimated in step ST104.

[0103] Next, the behavior identification function 157 identifies a recommended behavior to be recommended to the managed user (step ST107). For example, the behavior identification function 157 identifies, as a recommended behavior, a behavior whose executable level calculated in step ST103 is close to the executable range identified in step ST105 from the behaviors within the range of behaviors estimated in step ST106 to be able to reach the goal state.

[0104] Next, the presentation function 158 generates a presentation screen for presenting the identified recommended action to the user (step ST108). For example, the presentation function 158 generates a presentation screen for presenting the graph shown in FIG. 5 to the user.

[0105] Next, the display control function 159 performs control to display the presentation screen (step ST109), and ends this process. For example, the display control function 159 performs control to display the presentation screen generated in step ST108 on the display of the user terminal 203 via the NW interface 110.

[0106] As described above, the medical support device 100 according to this embodiment acquires the current state information of the managed person, sets a target state, and presents recommended action information indicating actions recommended for the managed person based on the current state information and the target state.

[0107] As a result, the medical support device 100 according to this embodiment can present the user with recommended actions to bring the managed person closer to the goal state, depending on the current state of the managed person. Therefore, the medical support device 100 according to this embodiment can reduce the possibility of presenting the user with actions that are not expected to be effective in alleviating the current symptoms of the managed person, or actions that are difficult to perform in the managed person's current living environment. In other words, the medical support device 100 according to this embodiment can present the subject with actions that are easy to perform and that are effective in alleviating the symptoms of lifestyle-related diseases and maintaining and improving physical condition.

[0108] Furthermore, the medical support device 100 according to this embodiment calculates, based on the current state information, a feasibility level indicating the degree of difficulty for the managed person to perform each of a plurality of actions, identifies candidate actions from among the plurality of actions for achieving the set target state, and presents recommended action information based on the feasibility level and the candidate actions.

[0109] As a result, the medical support device 100 according to this embodiment can present, for example, actions with a low feasibility level (actions that are easy for the managed person to perform) from among actions that can achieve the goal state.

[0110] Furthermore, the medical support device 100 according to this embodiment acquires current status information including information obtained by monitoring the person to be managed, such as pulse, respiration, blood pressure, body temperature, etc. This allows the medical support device 100 according to this embodiment to present, for example, recommended actions that match the information obtained by monitoring the person to be managed.

[0111] Furthermore, the medical support device 100 according to this embodiment acquires current status information including information obtained by questioning the person to be managed. As a result, the medical support device 100 according to this embodiment can present recommended actions that match, for example, lifestyle information about the life of the person to be managed obtained by questioning the person to be managed and attribute information representing the attributes of the person to be managed.

[0112] (Second embodiment) In the first embodiment, a form in which recommended actions are presented to the user has been described. In the second embodiment, a form in which barrier information regarding barriers that prevent the execution of an action is presented to the user in addition to the recommended actions will be described.

[0113] In the following, differences from the above-described embodiment will be mainly described, and detailed descriptions of commonalities with the contents already described will be omitted. Furthermore, each embodiment described below may be implemented individually or in appropriate combination.

[0114] First, the configuration of a medical support device 100a according to the second embodiment will be described. Fig. 7 is a diagram showing an example of the configuration of the medical support device 100a according to the second embodiment. As shown in Fig. 7, the medical support device 100a according to the second embodiment has substantially the same configuration as the medical support device 100 according to the first embodiment shown in Fig. 2, but differs from the first embodiment in that it includes a memory 120a and a processing circuit 150a.

[0115] The memory 120a stores a feasibility level calculation model 121, an effect estimation model 122, and an action list 123, as well as a barrier identification model 124 and correspondence information 125.

[0116] The barrier identification model 124 is a generation AI that generates sentences about barriers that prevent the managed person from performing an action. For example, the barrier identification model 124 is an LLM. As an example, the barrier identification model 124 generates sentences about barriers that prevent the managed person from performing an action in response to input of current state information and text describing the name of the action.

[0117] For example, barriers include time, concentration, mood, tools / environment, and meaning. Specifically, time refers to not being able to perform an action because there is no time. Concentration refers to not being able to perform an action because there is too much noise. Mood refers to not being able to perform an action because there is too much stress. Tools / environment refers to not being able to perform an action because there is not what is needed. Meaning refers to not being able to perform an action because there is no goal or enjoyment.

[0118] The correspondence information 125 is information for specifying a barrier countermeasure for reducing a barrier. The correspondence information 125 is information that associates an action, a barrier, and a solution (barrier countermeasure).

[0119] Here, Fig. 8 is a diagram showing an example of the data configuration of the correspondence information 125 according to the second embodiment. As shown in Fig. 8, this is information that associates the type of activity, the barrier, and the solution. The first line in Fig. 8 indicates that when the type of activity is "walking" and the barrier is "time," the solution is "XX."

[0120] In this embodiment, if the barrier is time, solutions to the barrier are set such as suggestions for actions that can be incorporated into the lifestyle of the person being managed, suggestions for short-term actions that can be carried out in the person's free time, suggestions for actions to create time, etc.

[0121] Furthermore, if the barrier is concentration, solutions to the barrier may include presenting information that it would be more effective for the person being managed to increase the frequency of other types of behavior, or presenting information about behavior that is considered unnecessary for improving the condition.

[0122] Furthermore, if the barrier is mood, a solution to the barrier may be set to suggest a method for relieving stress for the person to be managed.

[0123] Furthermore, if the barrier is a tool or environment, solutions to the barrier will include suggestions to encourage the acquisition of tools (preparing the environment), or suggestions for actions that have the same effect but do not require the tool (environment).

[0124] If the barrier is significance, solutions to the barrier may include providing information to help the person being managed reaffirm their purpose, offering incentives or penalties, etc. In this case, an incentive may be an increase in the discount rate on the usage fee for the affiliated service, etc., and a penalty may be the suspension of the discount on the usage fee for the affiliated service, etc.

[0125] The processing circuit 150a includes a learning function 151a, an acquisition function 152, a setting function 153, a calculation function 154, an estimation function 155, a range identification function 156, a behavior identification function 157, a presentation function 158a, a display control function 159, a barrier identification function 160, and a correspondence identification function 161. The acquisition function 152, the setting function 153, the calculation function 154, the estimation function 155, the range identification function 156, the behavior identification function 157, and the display control function 159 are the same as those in FIG. 2, and therefore their explanations will be omitted.

[0126] The learning function 151a learns the barrier identification model 124. The following describes the learning of the barrier identification model 124. For example, the learning function 151a collects text describing barriers that hinder the behavior registered in the behavior list 123.

[0127] As an example, the learning function 151a acquires an electronic medical record from the electronic medical record system 202 via the NW interface 110. The learning function 151a extracts text describing barriers to specific behavior and text describing current status information from the acquired electronic medical record.

[0128] The learning function 151a generates, as input training data, a set of text describing the name of a specific behavior type and text describing the extracted current state information. Furthermore, based on the text describing the extracted barriers to the specific behavior, the learning function 151a classifies the barriers into categories including at least one of time, concentration, mood, tools / environment, and significance. The learning function 151a generates text data describing the classification results as output training data.

[0129] For example, if a specific behavior is detected as being performed but for a short period of time, or if the specific behavior is only performed to the extent possible within a person's daily routine, such as commuting, and the person's daily life is too busy, the learning function 151a classifies the barrier as time. Furthermore, if a specific behavior is detected as being performed but infrequently, the learning function 151a classifies the barrier as concentration.

[0130] Furthermore, the learning function 151a classifies the barrier as a tool / environment when the tool (e.g., a bicycle in cycling) or environment (e.g., a pool in swimming) required for the behavior is not present. Furthermore, the learning function 151a classifies the barrier as a mood when the specific behavior is performed only to the extent that it is feasible within the context of daily life and the stress value is high.

[0131] In addition, the learning function 151a classifies the barrier as significance if the specific behavior is performed only to the extent that it is feasible within the context of the person's life, the person's life is not too busy, and the stress level is not high.

[0132] The learning function 151a may output the text input by the user as the barrier corresponding to the input training data as the output training data. In this case, the user may ask the person to be managed about the barrier and input the barrier according to the result of the questioning.

[0133] The learning function 151a also generates a set of the generated input training data and the generated output training data as a training data set. The learning function 151a uses the generated training data set to train the barrier identification model 124 by a well-known deep learning technique.

[0134] The barrier identification function 160 identifies barriers to behavior. For example, the barrier identification function 160 inputs, into the barrier identification model 124, text describing the names of behavior types for which no words related to behavior were extracted when the range identification function 156 identified the range of behavior (or types of behavior that were detected but performed infrequently) and the current state information acquired by the acquisition function 152.

[0135] Hereinafter, the type of behavior for which no execution-related words were extracted will also be referred to as the type of behavior for which no execution was detected.Furthermore, the type of behavior for which execution-related words were detected but which was performed infrequently will also be referred to as the type of behavior for which execution was infrequent.

[0136] The barrier identification function 160 identifies barriers for the types of behavior for which no execution was detected and for which execution is infrequent when the range identification function 156 identifies the range of execution, in accordance with the output of the barrier identification model 124.

[0137] Here, Fig. 9 is a diagram illustrating an example of the results of identifying barriers to behavior according to the second embodiment. Fig. 9 shows an example of the results of identifying barriers to eating, cycling, and swimming for a specific person to be managed. "Time," "Concentration," and "Tools and environment" in Fig. 9 represent barriers to behavior.

[0138] In the example of Fig. 9, although the scope identification function 156 has extracted words related to the execution of "eating," the frequency of these words is low, so the barrier identification function 160 inputs the text describing "eating" and the current state information acquired by the acquisition function 152 into the barrier identification model 124. Following the output of the barrier identification model 124, the barrier identification function 160 identifies the barriers to eating as "time" and "concentration."

[0139] 9, since no words related to execution have been extracted for swimming by the range identification function 156, the barrier identification function 160 inputs the text describing swimming and the current state information acquired by the acquisition function 152 into the barrier identification model 124. According to the output of the barrier identification model 124, the barrier identification function 160 identifies the barrier to swimming as "tools and environment."

[0140] 9, since no words related to execution have been extracted for cycling by the range identification function 156, the barrier identification function 160 inputs the text describing cycling and the current state information acquired by the acquisition function 152 into the barrier identification model 124. Following the output of the barrier identification model 124, the barrier identification function 160 identifies the barrier to cycling as "tools and environment."

[0141] The correspondence identification function 161 identifies a barrier correspondence that corresponds to a barrier to a behavior. For example, the correspondence identification function 161 refers to the correspondence information 125 stored in the memory 120a and identifies a solution that corresponds to the barrier identified by the barrier identification function 160 as the barrier correspondence.

[0142] The presentation function 158a generates a presentation screen that presents information representing barriers and barrier solutions. For example, the presentation function 158a generates a presentation screen that includes the graph of Fig. 9 and information representing solutions to barriers for eating, cycling, and swimming. Note that the presentation function 158a may generate a presentation screen that displays solutions by, for example, clicking on the display of a barrier corresponding to an activity.

[0143] Next, the processing executed by the medical support device 100a will be described. Fig. 10 is a flowchart showing an example of the processing executed by the medical support device 100a according to the second embodiment. Steps ST201 to ST207 are the same as steps ST101 to ST107 in Fig. 6, and therefore their description will be omitted.

[0144] After step ST207, the barrier identification function 160 identifies a barrier to the behavior (step ST208).

[0145] For example, when identifying the execution range in step ST205, the barrier identification function 160 inputs text describing the name of the type of behavior for which execution has not been detected or the name of the type of behavior for which execution is infrequent, and the current state information acquired in step ST201, into the barrier identification model 124. The barrier identification function 160 identifies a barrier for the type of behavior in question according to the output of the barrier identification model 124.

[0146] Next, the correspondence identification function 161 identifies a barrier correspondence for the barrier to the behavior (step ST209). For example, the correspondence identification function 161 refers to the correspondence information 125 stored in the memory 120a, and identifies, as a barrier correspondence, a solution that corresponds to the type of behavior whose execution was not detected in step ST205 (or the type of behavior whose execution is infrequent) and the barrier to the behavior identified in step ST208.

[0147] The presentation function 158 generates a presentation screen for presenting the identified recommended behavior to the user and a presentation screen for presenting the identified behavioral barrier and the corresponding barrier to the behavioral barrier to the user (step ST210). For example, the presentation function 158 generates a presentation screen for presenting the graph shown in Fig. 5 and the graph shown in Fig. 9 and the corresponding solutions to the behavioral barrier in the graph to the user.

[0148] Step ST211 is similar to step ST109 in FIG. 6, and therefore a description thereof will be omitted.

[0149] As described above, the medical support device 100a according to this embodiment identifies barriers that hinder the execution of a type of behavior for which execution has not been detected, based on the acquired current state information, and presents information about the barriers along with recommended behavior.

[0150] As a result, the medical support device 100a according to this embodiment can present to the user, for the type of behavior that the managed person has not performed, information on the barriers to that type of behavior. The user who has received the presented information on the barriers can understand, for example, the reasons why the managed person has not been able to perform a behavior that is thought to be effective in improving the condition but that the managed person has not been able to perform.

[0151] Furthermore, the medical support device 100a according to this embodiment identifies a barrier countermeasure to reduce the barrier for the identified barrier type, and presents information about the identified barrier countermeasure together with information about the barrier.

[0152] As a result, the medical support device 100a according to this embodiment can present the user with a solution to reduce barriers to a type of behavior that the managed person has not performed, for example. The user who receives the solution can then provide the managed person with appropriate advice to take an action that is effective in improving their condition.

[0153] The above-described embodiment can be modified as needed by partially changing the configuration or functions of each device. Therefore, several modifications of the above-described embodiment will be described below as other embodiments. The following mainly focuses on differences from the above-described embodiment, and detailed descriptions of commonalities with the content already described will be omitted. The modifications described below may be implemented individually or in appropriate combination.

[0154] (Variation 1) In the first and second embodiments described above, the behavior identification function 157 identifies, as a recommended behavior, an behavior that is close to the goal state among the behaviors that are estimated to be capable of reaching the goal state. In this modification, a description will be given of a behavior in which a trained model trained by machine learning techniques including deep learning is used to identify a recommended behavior.

[0155] The learning function 151 of this modification learns a learned model for identifying a recommended action (hereinafter also referred to as an action identification model).

[0156] The behavior identification model is a generation AI that generates sentences regarding recommended actions to bring the state of the managed individual closer to the goal state. For example, the behavior identification model is an LLM. As an example, the behavior identification model generates sentences regarding recommended actions to bring the state of the managed individual closer to the goal state in response to input of current state information and text describing the goal state.

[0157] For example, the learning function 151 collects, from multiple managed persons, current state information acquired by the acquisition function 152 and text describing a target state set by the setting function 153 corresponding to the acquired current state information. The learning function 151 generates a set of the collected current state information and the text describing the target state as input training data.

[0158] Furthermore, the learning function 151 collects text describing information about the recommended behavior identified by the behavior identification function 15 based on the current state information and the goal state for each of the multiple managed individuals. The learning function 151 generates the text describing the collected information about the recommended behavior as output training data.

[0159] The learning function 151 also generates a set of the generated input training data and output training data (recommended actions identified from the current state information and the goal state) corresponding to the contents of the input training data (the current state and the goal state) as a training dataset. The learning function 151 uses the generated training dataset to learn an action identification model using a known deep learning technique.

[0160] The behavior identification function 157 of this modification identifies a recommended behavior using a trained model. For example, the behavior identification function 157 inputs the current state information acquired by the acquisition function 152 and text describing the goal state set by the setting function 153 into the behavior identification model. The behavior identification function 157 identifies a recommended behavior according to the output of the behavior identification model.

[0161] According to this modification, after the behavior specification model is generated, it is possible to present recommended behaviors without performing processes such as specifying an execution range by the range specification function 156 or specifying a range of behaviors that are estimated to enable reaching a goal state by the behavior specification function 157. This reduces the processing load on the medical support device 100.

[0162] (Variation 2) In the second embodiment described above, a configuration has been described in which the correspondence identification function 161 identifies a barrier correspondence (solution) for a barrier to a behavior using the correspondence information 125. In this modification, a configuration will be described in which a trained model trained by machine learning technology including deep learning is used to identify a barrier correspondence.

[0163] The learning function 151a of this modification learns a trained model for identifying barrier correspondence (hereinafter also referred to as correspondence identification model).

[0164] The response identification model is a generative AI that generates sentences related to barrier responses that reduce barriers to behavior. For example, the response identification model is an LLM. As an example, the response identification model generates sentences related to barrier responses that reduce barriers to behavior in response to inputs of text describing the type of behavior, text describing the barriers to the behavior, and current state information.

[0165] For example, the learning function 151a collects, from multiple managed persons, current state information acquired by the acquisition function 152 and text describing barriers to behavior identified by the barrier identification function 160. The learning function 151a generates, as input training data, a set of the collected current state information, text describing barriers to behavior, and text describing the type of behavior corresponding to the barriers to the behavior.

[0166] Furthermore, the learning function 151a collects text describing information about barrier responses identified by the response identification function 161 for each of the multiple managed individuals, based on the current status information, the type of behavior, the behavioral barriers, and the response information 125. The learning function 151a generates the text describing the collected information about barrier responses as output training data.

[0167] The learning function 151a may extract words related to actions actually taken to reduce barriers to behavior from electronic medical records, etc., and generate text based on the extracted content as output training data. In this case, the process of identifying barrier responses by the response identification function 161 may not be performed.

[0168] Furthermore, the learning function 151a generates a set of the generated input teacher data and output teacher data (current state information, type of action, barrier to action, and barrier correspondence identified from correspondence information 125) corresponding to the contents of the input teacher data (current state, type of action, barrier to action), as a learning dataset. The learning function 151a uses the generated learning dataset to learn a barrier identification model by known deep learning technology.

[0169] The correspondence identification function 161 of this modification identifies a recommended behavior using a trained model. For example, the correspondence identification function 161 inputs the current state information acquired by the acquisition function 152, text describing the barriers to the behavior identified by the barrier identification function 160, and text describing the type of behavior corresponding to the barrier to the behavior into the behavior identification model. The behavior identification function 157 identifies the barrier correspondence according to the output of the behavior identification model.

[0170] According to this modification, for example, the process of searching the correspondence information 125 is not necessary, so that it is expected that the processing load on the medical support device 100a will be reduced in situations where the amount of information registered in the correspondence information 125 becomes enormous.

[0171] (Variation 3) In the above-described first and second embodiments, the functions are realized by using a processor installed in the medical support device 100 (100a). However, the processing circuit 150 (150a) of the medical support device 100 (100a) may realize the functions by using a processor of an external device connected via a network N.

[0172] In this modified example, for example, the processing circuit 150 (150a) reads and executes a program corresponding to each function from the memory 120 (120a), and realizes each function shown in FIG. 2 (or FIG. 7) by using a group of servers (cloud) connected to the medical support device 100 (100a) via the network N as a computational resource.

[0173] The various data handled in this specification are typically digital data.

[0174] According to at least one of the embodiments described above, it is possible to present to a subject actions that are effective in alleviating symptoms of lifestyle-related diseases and maintaining and improving physical condition, and that are easy for the subject to carry out.

[0175] Although several embodiments have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, modifications, and combinations of embodiments can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, as well as within the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0176] 100 Medical support equipment 110 Network Interface 120 memory 130 Input Interface 140 Display 150 Processing Circuit 151 Learning Function 152 Acquisition Function 153 Setting Function 154 Calculation Function 155 Guessing Function 156 Range Identification Function 157 Behavioral Identification Function 158 Presentation function 159 Display Control Function 160 Barrier Identification Function 161 Supported specific functions 200 Medical information storage device 201 Medical Interview System 202 Electronic Medical Record System S Medical Support System

Claims

1. an acquisition unit that acquires current state information representing the current state of the subject; a setting unit that sets a target state representing a target state of the subject; a presentation unit that presents recommended behavior information representing a behavior recommended for the subject based on the current state information and the goal state; An information processing device comprising:

2. a calculation unit that calculates, based on the current state information, an execution difficulty level indicating a degree of difficulty when the subject executes each of a plurality of actions; a behavior identification unit that identifies, from among a plurality of behaviors, a behavior candidate that represents a candidate behavior for achieving the goal state represented by the goal state; Further provided with the presenting unit presents the recommended action information based on the execution difficulty and the candidate actions. The information processing device according to claim 1 .

3. the acquisition unit acquires the current state information including various monitor information obtained by monitoring the subject; a range specifying unit that specifies an action range indicating a range of an action being performed by the subject from the plurality of actions based on the monitoring information; the presenting unit presents, as the recommended behavior information, behaviors outside the execution range based on the execution difficulty level. The information processing device according to claim 2 .

4. the acquiring unit acquires the current condition information including interview information obtained by interviewing the subject, the range specification unit specifies the execution range from the plurality of actions based on the medical interview information; The information processing device according to claim 3 .

5. the calculation unit inputs the action and the current state information into a first trained model that has learned a relationship between the action, the current state information, and the execution difficulty by machine learning, and calculates the execution difficulty based on an output of the first trained model; the behavior identification unit inputs the behavior and the current state information into a second trained model that has learned, by machine learning, a relationship between the behavior, the current state information, and an improvement effect indicating the degree of improvement of the subject's condition caused by performing the behavior; estimates the improvement effect caused by the subject performing the behavior based on an output of the second trained model; and identifies the candidate behavior based on the goal state and the estimated improvement effect. The information processing device according to claim 2 .

6. a barrier identification unit that identifies a barrier to the execution of the action outside the identified execution range based on the current state information; the presentation unit presents the barrier together with the recommended behavior information. The information processing device according to claim 1 .

7. a correspondence specifying unit that specifies a barrier correspondence for reducing the barrier based on the current state information and the barrier, the presentation unit presents the barrier correspondence together with the barrier; The information processing device according to claim 6 .

8. an acquisition step of acquiring current state information representing a current state of the subject; a setting step of setting a target state representing a target state of the subject; a presentation step of presenting recommended behavior information representing a behavior recommended to the subject based on the current state information and the goal state; An information processing method including:

9. On the computer, an acquisition step of acquiring current state information representing a current state of the subject; a setting step of setting a target state representing a target state of the subject; a presentation step of presenting recommended behavior information representing a behavior recommended to the subject based on the current state information and the goal state; A program that executes the following.

Images