Planning device, planning method, and program

Abstract

To enable those implementing a plan that involves gradually changing a combination of quantitative values ​​to verify the results of the plan's execution. [Solution] The planning device obtains feasibility evaluation index values ​​that quantitatively show the ease of realizing each state for each state identified using the values ​​of one or more items that correlate with the health status evaluation index value, obtains the health status evaluation index value for each state, and uses evaluation values ​​for paths consisting of one or more state transitions, calculated using evaluation values ​​for each state transition, where the evaluation is given as an evaluation for a single state transition from the source state to the destination state, with the evaluation being given as the easier it is to realize the destination state, and the evaluation being given as the better the health status evaluation index value in the destination state is compared to the health status evaluation index value in the source state, to search for a path from the initial state to the target state in the plan.

Description

[Technical Field]

[0001] The present invention relates to a planning device, a planning method, and a program. [Background technology]

[0002] To improve an individual's health, it is conceivable to develop a plan that involves gradually changing a combination of quantitative values, such as BMI (Body Mass Index) and blood pressure. When developing such a plan, it is advisable to ensure that the combination of quantitative values ​​shown in the plan is highly feasible in order to make it easy to implement.

[0003] For example, the health improvement path search device described in Patent Document 1 treats the variable space of multiple explanatory variables selected from variables that are human measurement values ​​measured by health checkups, etc., as a graph divided into a grid, and obtains a health improvement path from among the paths that connect the grid points as nodes. Specifically, the health improvement path search device identifies paths that transition each measurement target value, starting from the current values ​​of the multiple explanatory variables, and designates paths in which the value of the health index at the endpoint is improved compared to the health index at the current value as candidate paths. Then, the health improvement path search device identifies the path among the candidate paths that has the maximum product of the probability of existence of each measurement target value within the path (the likelihood of existence for each combination of the value of the explanatory variable and the value of the health index), as the health improvement path. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] International Publication No. 2022 / 085785 [Overview of the project] [Problems that the invention aims to solve]

[0005] If individuals implementing a plan that involves gradually changing a combination of quantitative values ​​can confirm the results of their plan execution, it is expected that this will motivate them to continue the plan. This motivation to continue the plan is expected to lead to its actual implementation.

[0006] One example of the purpose of this disclosure is to provide a planning device, a planning method, and a program that can solve the problems described above. [Means for solving the problem]

[0007] According to a first aspect of this disclosure, the planning device includes: feasibility evaluation index value acquisition means for acquiring feasibility evaluation index values ​​that quantitatively indicate the ease of realizing each state, for each state identified using the values ​​of one or more items correlated with the health status evaluation index value; health status evaluation index value acquisition means for acquiring the health status evaluation index value for each state; and path search means for searching a path from an initial state to a target state in a plan using evaluation values ​​for each state transition, which are calculated using evaluation values ​​for each state transition, where the evaluation indicates a better rating the easier it is to realize the target state, and the evaluation indicates a better rating the health status evaluation index value in the target state shows a better rating than the health status evaluation index value in the source state.

[0008] According to a second aspect of this disclosure, the planning method includes a computer obtaining feasibility evaluation index values ​​that quantitatively indicate an evaluation of how easily each of the states is realized, for each state identified using the values ​​of one or more items correlated with a health status evaluation index value; obtaining the health status evaluation index value for each of the states; and using evaluation values ​​for one or more states, calculated using evaluation values ​​for each state transition, where the evaluation for a single state transition from the source state to the destination state is better the easier it is to realize the destination state, and the evaluation is better the better the health status evaluation index value in the destination state is compared to the health status evaluation index value in the source state. The computer then searches for a path from an initial state to a target state in the plan using evaluation values ​​for one or more states transitions.

[0009] According to a third aspect of this disclosure, the program causes a computer to perform the following actions: obtain feasibility evaluation index values ​​that quantitatively indicate the ease of realizing each state, for each state identified using the values ​​of one or more items correlated with the health status evaluation index value; obtain the health status evaluation index value for each state; and search for a path from an initial state to a target state in a plan using evaluation values ​​for each state transition, where the evaluation indicates a better rating the easier it is to realize the target state, and the evaluation indicates a better rating the health status evaluation index value in the target state shows a better rating than the health status evaluation index value in the source state. [Effects of the Invention]

[0010] According to one aspect of this disclosure, it is expected that a person who implements a plan that involves gradually changing a combination of quantitative values ​​will be able to confirm the results of the plan's implementation. [Brief explanation of the drawing]

[0011] [Figure 1]This figure shows an example of the configuration of a planning device according to at least one embodiment. [Figure 2] This figure shows an example of displaying feasibility evaluation values ​​by a display unit according to at least one embodiment. [Figure 3] This figure shows an example of displaying health status evaluation index values ​​by a display unit according to at least one embodiment. [Figure 4] This figure shows an example of displaying evaluation index values ​​for each state using a display unit according to at least one embodiment. [Figure 5] This figure shows an example of a feasibility assessment index model used by a feasibility assessment index value acquisition unit according to at least one embodiment. [Figure 6] This figure shows an example of the procedure for generating and outputting a plan using a planning device according to at least one embodiment. [Figure 7] This figure shows an example of the procedure for a pathfinding process performed by a pathfinding unit according to at least one embodiment. [Figure 8] This figure shows an example of the procedure for a process in which a route search unit according to at least one embodiment performs initial settings for route searching. [Figure 9] This figure shows an example of the procedure for a planning device to generate and output a plan when a target setting unit sets a target state according to at least one embodiment, and then a path search unit performs a path search. [Figure 10] This figure shows an example of the procedure for the planning device to generate and output a plan when extending the search range according to at least one embodiment. [Figure 11] This figure shows an example of a procedure for a planning device according to at least one embodiment that generates and outputs a plan spanning multiple time periods. [Figure 12] This figure shows an example of a feasibility assessment index model used by the feasibility assessment index value acquisition unit when a planning device according to at least one embodiment generates a plan spanning multiple periods. [Figure 13] This figure shows an example of displaying a plan spanning multiple periods using a display unit according to at least one embodiment. [Figure 14] This figure shows a first example of the display of period-specific data by a display unit according to at least one embodiment. [Figure 15] This figure shows a second example of the display of period-specific data by a display unit according to at least one embodiment. [Figure 16] This figure shows an example of the configuration of a planning device according to at least one embodiment. [Figure 17] This figure shows an example of a processing procedure in a planning method according to at least one embodiment. [Figure 18] This figure shows an example of a computer configuration according to at least one embodiment. [Modes for carrying out the invention]

[0012] The embodiments will be described below with reference to the drawings. In the following, characters with a circumflex symbol may be indicated by adding a "^" after the character. For example, the circumflex-equipped X will also be written as X^. <First Embodiment> Figure 1 shows an example of the configuration of a planning device according to at least one embodiment. In the configuration shown in Figure 1, the planning device 100 includes a communication unit 110, a display unit 120, an operation input unit 130, a storage unit 180, and a processing unit 190. The processing unit 190 includes a data acquisition unit 191, a health status evaluation index value acquisition unit 192, a feasibility evaluation index value acquisition unit 193, a goal setting unit 194, and a route search unit 195.

[0013] The planning device 100 generates a plan to improve health status. Specifically, the planning device 100 plans changes in values ​​correlated with the health status evaluation index, which is an evaluation index for health status, so that the value of the health status evaluation index shows a better evaluation. The health status evaluation index can be understood as an index that shows an evaluation of a certain aspect of health status (the state of the body, mind, or a combination thereof).

[0014] The planning device 100 may be configured using a computer such as a personal computer (PC) or a workstation (WS). The plan generated by the planning device 100 can also be called a health promotion plan. The act of the planning device 100 generating a plan can also be understood as formulating a plan. The plan generated by the planning device 100 can also be simply referred to as a plan.

[0015] A medical professional, such as a physician in charge of a health checkup, may use the planning device 100 to obtain a plan. The medical professional may then present the plan generated by the planning device 100 to the person who will be the subject of the plan (the person who will carry out the plan). Alternatively, the medical professional may use the plan generated by the planning device 100 as a reference to formulate their own plan, and then present the plan they themselves formulated to the person who will be the subject of the plan. Alternatively, a person who wishes to improve their health condition (the subject of the plan) may use the planning device 100 to obtain a plan. The target group of the plan is also simply referred to as the target group.

[0016] The communication unit 110 communicates with other devices. For example, the communication unit 110 may receive data necessary for generating a plan, such as the health checkup results of the subject, from other devices. Alternatively, the communication unit 110 may transmit the plan generated by the planning device 100 to other devices.

[0017] The display unit 120 includes a display screen such as a liquid crystal panel or an LED (Light Emitting Diode) panel, and displays various images. For example, the display unit 120 may display the plan generated by the planning device 100. The operation input unit 130 includes, for example, input devices such as a keyboard and a mouse, and accepts user operations. For example, the operation input unit 130 may be configured to accept input operations for various settings related to the generation of the plan, such as the numerical range to be planned.

[0018] The memory unit 180 stores various types of data. For example, the memory unit 180 may store various models, such as a model for health assessment indicators (a model for calculating health assessment indicator values). The memory unit 180 is configured using the memory devices provided by the planning device 100.

[0019] The processing unit 190 controls various parts of the planning device 100 to perform various processes. The functions of the processing unit 190 are performed, for example, by the CPU (Central Processing Unit) of the planning device 100 reading a program from the storage unit 180 and executing it.

[0020] The data acquisition unit 191 acquires various data for the planning device 100 to generate a plan. For example, the data acquisition unit 191 may acquire the health checkup results of the subject from other devices via the communication unit 110.

[0021] The health status evaluation index value acquisition unit 192 acquires health status evaluation index values ​​for each state within the range set as the target of the plan. The health status evaluation index value acquisition unit 192 is an example of a means for acquiring health status evaluation index values. The health status evaluation index acquisition unit 192 may calculate health status evaluation index values ​​for each state. For example, the health status evaluation index acquisition unit 192 may calculate health status evaluation index values ​​using a trained model of health status evaluation index.

[0022] Alternatively, the health status evaluation index value acquisition unit 192 may acquire health status evaluation index values ​​determined for each state. For example, the health status evaluation index value acquisition unit 192 may read health status evaluation values ​​from the storage unit 180, or it may acquire health status evaluation index values ​​from other devices via the communication unit 110.

[0023] The state referred to here (the state in the plan) is a health state (the state of the body, mind, or a combination thereof). In the plan generated by the planning device 100, discrete states are used, and the state is identified using one or more items that correlate with the health status evaluation index value. The items used to identify a state are those whose values ​​will be directly changed or maintained in the plan. That is, the plan presents target values ​​for the items used to identify the state, and the person takes action to achieve those target values. This action may include simply living one's life.

[0024] Items used to identify a condition are also called condition identification items. Preferably, condition identification items are items that the subject can relatively easily know the value of, such as items measured in a health checkup, or items that the subject can measure or calculate themselves. A continuous value may be used as a state-specific item, and intervals obtained by dividing the state-specific item values ​​may be assigned to states. Furthermore, a representative value such as the median of an interval assigned to a state may be used as the state-specific item value for that state. For example, the representative value assigned to a state may be used to identify a state and to calculate health assessment index values ​​for that state. A state-specific item value that identifies a particular state is also called the state-specific item value for that state.

[0025] The processing unit 190 may also set status identification items. For example, the processing unit 190 may select a predetermined number of items from among the items measured in the health checkup that have the strongest correlation with the health status evaluation index value as status identification items. Alternatively, the data acquisition unit 191 may set the status identification items, such as by reading the health status evaluation index items and status identification items from the storage unit 180.

[0026] The scope of the plan may be defined as a predetermined range based on the initial state in the plan. For example, the scope of the plan may be defined as the range that can be reached within a predetermined number of state transitions from the initial state in the plan. The scope of the plan (the range set as the target of the plan) is also called the search range. A single state transition can be defined as a transition (movement) from one state to another adjacent state. Adjacent states may have adjacent or identical assigned intervals (numerical ranges) for any of their state-specific items, and may not be identical states. Alternatively, adjacent states may have adjacent assigned intervals for any one of their state-specific items, while having the same assigned intervals for the other state-specific items.

[0027] The processing unit 190 may set the search range. Alternatively, the data acquisition unit 191 may set the search range by reading the search range information from the storage unit 180. The initial state in the plan may be the state of the subject at the time the plan was created. The initial state in the plan may also be called the starting state. The initial state in the plan may also simply be referred to as the initial state or starting state.

[0028] The following explanation uses an example where the health assessment index is the risk of developing diabetes, and weight and blood glucose are used as condition identification items. However, the items included in the health assessment index are not limited to specific items, but can be various items that allow for the determination of a state-specific item and the measurement or calculation of a health assessment item value. The state-specific item is not limited to specific items, but can be various items that correlate with the health assessment index. The number of state-specific items is also not limited to a specific number, but can be one or more.

[0029] The risk of developing the disease referred to here may represent the probability of developing the target disease within a predetermined period. For example, the risk of developing the disease may be the probability of developing diabetes in the next 10 years, or the probability of developing diabetes in the next 3 years. The diseases targeted for assessment of risk of developing the disease are not limited to diabetes. For example, the diseases targeted for assessment of risk of developing the disease may be heart disease or brain disease, but are not limited to these.

[0030] For example, the health status evaluation index value acquisition unit 192 may input health checkup data into a model that predicts the risk of developing the disease and calculate the risk of developing the disease. The model that predicts the risk of developing the disease is also called a disease risk prediction model, or simply a prediction model. The predictive model may be obtained through machine learning. For example, inputting health checkup data of a subject into the predictive model may allow for the calculation of the probability of developing a target disease within three years.

[0031] This section describes predictive models for forecasting the risk of developing diabetes. Here, we will explain an example of how to construct a predictive model that calculates the probability of developing diabetes within three years from a base year (the year used as the base year for calculating the risk of developing diabetes).

[0032] The data used to build the predictive model is (X1,Y1), (X2,Y2), ..., (X N ,Y N ) Let the data be (X1,Y1), (X2,Y2), ..., (X N ,Y N Each of these is also called training data. Data (X1,Y1), (X2,Y2), ..., (X N ,Y N The set of these elements is also called the training dataset.

[0033] N is an integer N≧1 that represents the number of people included in the training dataset (the people from whom the training data is measured). These people are also referred to as the subjects of the health examination. The subjects of the health examination will be identified as 1, ..., N, and the training data (X n, Y n The examinee who is the measurement source of ) is also denoted as examinee n. Here, n is an integer such that 1 ≤ n ≤ N.

[0034] X n represents the health examination data of examinee n for the reference year. Assume that the health examination data includes data on M health examination items. M is an integer indicating the number of health examination items (number of items), M ≥ 1. The health examination items are also denoted as health examination item 1, health examination item 2,..., health examination item M. Assume that among the M health examination items, there are items for specifying the state. The health examination items may be, for example, weight, height, blood sugar, blood pressure, HDL cholesterol, LDL cholesterol, etc. The measured value of health examination item j of examinee n is denoted as X n,j Here, j is an integer such that 1 ≤ j ≤ M.

[0035] Assume that none of the N examinees have developed diabetes at the time of the reference year. Y n represents a flag indicating whether diabetes has developed within 3 years from the reference year. Y n = 1 indicates the case of development, and Y n = 0 indicates the case of no development. The flag indicating whether diabetes has developed within 3 years from the reference year is also referred to as the development flag. P(Y = 1) represents the probability of development that a person will develop diabetes within 3 years from the reference year.

[0036] The prediction model is constructed, for example, using the training data set {(X1, Y1), (X2, Y2),..., (X N , Y N )} for the reference year. Consider constructing a prediction model that receives the input of the health examination data of the subject and outputs the probability of development that the individual (subject) will develop diabetes within 3 years from the reference year.

[0037] The logistic regression model is one example of a model capable of such inputs and outputs. However, any model capable of such inputs and outputs other than the logistic regression model may be used. The following section explains the case where a logistic regression model is used.

[0038] Let X be an M-dimensional explanatory variable corresponding to the health checkup data for the base year, and let Y be the dependent variable representing whether or not diabetes developed within 3 years from the base year. Given that W is an M-dimensional weight vector, the conditional probability P(Y=1|X;W) that Y=1 when the value of X (health checkup data) is given can be expressed as shown in equation (1).

[0039]

number

[0040] The superscript T indicates the transpose of a vector or matrix. The weight vector W corresponds to the parameters that are adjusted in machine learning. The conditional probability P(Y=0|X;W) that Y=0 given a value for X is given is given by equation (2).

[0041]

number

[0042] The training dataset {(X1,Y1),(X2,Y2),…,(X N ,Y N Given ), logistic regression searches for values ​​in the weight vector W to optimize (in this case, maximize) the value of the objective function shown in equation (3).

[0043]

number

[0044] X n and Y n These represent the health check data and onset flag values ​​for the patient n, respectively. The value of the objective function L(W) can be maximized, for example, using a method similar to the gradient method. Thus, given (X n ,Y n Given the data (n=1,…,N), the probability P(Y n |X n The method for finding the value of the weight vector W that maximizes the sum of probabilities when calculating W is known as the maximum likelihood estimation method.

[0045] Here, let W* be the value of the weight vector W that maximizes the objective function L(W). Using W* and the values ​​X of M health checkup items for the subject (the individual whose risk of developing diabetes we want to predict) in the base year, the probability of developing diabetes within 3 years from the base year can be calculated by P(Y=1|X;W*). As mentioned above, the M health checkup items are assumed to include values ​​for condition-specific items.

[0046] The feasibility evaluation index value acquisition unit 193 acquires the feasibility evaluation index value for each state within the search range. The feasibility evaluation index value acquisition unit 193 is an example of a means for acquiring feasibility evaluation index values. The feasibility assessment index referred to here is an index that quantitatively shows the ease with which a state can be achieved. For example, if weight and blood glucose are used as state-specific items, the feasibility assessment value would indicate the ease with which the combination of weight and blood glucose can be achieved.

[0047] The feasibility evaluation index acquisition unit 193 may calculate the feasibility evaluation index value for each state. For example, the feasibility evaluation index acquisition unit 193 may calculate the feasibility evaluation index value using a trained feasibility evaluation index model (a model for calculating feasibility evaluation index values).

[0048] Alternatively, the feasibility evaluation index value acquisition unit 193 may acquire feasibility evaluation index values ​​determined for each state. For example, the feasibility evaluation index value acquisition unit 193 may read feasibility evaluation index values ​​from the storage unit 180, or it may acquire feasibility evaluation index values ​​from other devices via the communication unit 110.

[0049] Feasibility assessment index values ​​may be determined using statistical data. For example, statistical data may be assigned to each state, and the feasibility assessment index value for each state may be calculated such that the more data assigned to a state, the better the feasibility assessment of that state becomes (the higher the feasibility). While it is possible to use the probability of a state occurring (the probability that the state will be realized) as an index for evaluating the feasibility of a certain state, the index is not limited to this.

[0050] Here, the number of state-specific items (number of items) is Q, and these are also referred to as state-specific item 1, state-specific item 2, ..., state-specific item Q. Q is an integer such that 1 ≤ Q ≤ M. The value of the state-specific item q (state-specific item value) is also denoted as X^q. Here, q is an integer between 1 and Q. The state item value X^ of a single person can be represented as a Q-dimensional vector, as shown in equation (4).

[0051]

number

[0052] When distinguishing between status-specific item values ​​for multiple people, they should be expressed as shown in formula (5).

[0053]

number

[0054] Here, n is an integer n≧1, and it represents an index for identifying a person. When weight and blood glucose are used as condition-specific items, the feasibility assessment value indicates how easily the combination of weight and blood glucose can be achieved. The following explains how to calculate the probability of each condition occurring, using the case where the condition-determining items are weight and blood glucose as an example. When the condition-determining items are weight and blood glucose, the condition-determining item value X^ for one person can be represented by a two-dimensional vector (X^1, X^2). X^1 represents the weight value. X^2 represents the blood glucose value.

[0055] The types (choices) of values ​​that the weight value X^1 can take are X^ i Let 1 (i=1,...,K), and the number of possible values ​​for blood glucose X^2 is X^ j If we set 2(j=1,…,L), then P(X^1=X^ i 1, X^2=X^ j 2) The weight value X^1 is X^ i In 1, the blood glucose level X^2 ​​is X^ j This represents the probability of the event occurring where the result is 2. K represents the number of possible values ​​for weight X^1 (the number of options for the weight X^1 value). L represents the number of possible values ​​for blood glucose X^2 (the number of options for the blood glucose X^2 value). Probability P(X^1=X^ i 1, X^2=X^ j 2) can also be called the probability of occurrence. Probability P(X^1=X^ i 1, X^2=X^ j Regarding (2), equation (6) holds true.

[0056]

number

[0057] Table 1 shows examples of the number of people for each combination of body weight and blood glucose level, given X^=(X^1,X^2).

[0058] [Table 1]

[0059] The value in each cell in Table 1 is the weight value X^ i 1. And, blood glucose level is X^ j This indicates the number of people in state 2. Table 1 shows an example of the results of classifying 211 individuals whose possible values ​​for weight X1 are 60, 62, 64, and 66, and whose possible values ​​for blood glucose X2 are 105, 110, 115, and 120. In the example in Table 1, K=4 and L=4. Also, N=211. In the case of Table 1, the weight value X^1 is X^ i Take 1, and the blood glucose level X^2 ​​is X^ j Table 2 shows the probabilities of taking a 2.

[0060] [Table 2]

[0061] The probabilities shown in each cell of Table 2 are calculated based on the weight value shown in each cell of Table 1, where X^ i 1, and blood glucose level is X^ j The number of people who are 2 can be calculated by dividing that number by the total number of people N. For the probabilities shown in Table 2, equation (7) holds true.

[0062]

number

[0063] Here, we explained the case where the state-specific item values ​​are discrete, but the probability of occurrence can be calculated similarly even when the state-specific item values ​​are continuous.

[0064] Health status assessment index values ​​may also be determined using statistical data. For example, statistical data may be assigned to each state, and the health status assessment index values ​​for each state may be determined based on the state of the health status assessment index items in the assigned data. If the health status assessment index is the risk of developing diabetes, and the state-specific items are weight and blood glucose, the data may be assigned to each state based on weight and blood glucose values, and the ratio of the number of data points that have developed diabetes (the number of people who have developed diabetes) out of the number of data points assigned to a state (the number of people assigned to a state) may be used as the health status assessment index value for that state. However, the method for determining the health status index values ​​is not limited to a specific method.

[0065] The feasibility evaluation index value acquisition unit 193 may acquire feasibility evaluation index values ​​for the states necessary for generating the plan when the planning device 100 generates the plan. Feasibility evaluation index values ​​are not necessarily required for determining the target state in the plan. For this reason, the feasibility evaluation index value acquisition unit 193 does not necessarily need to acquire feasibility evaluation index values ​​for all states within the search range. The target state in a plan is also simply referred to as the target state.

[0066] The target setting unit 194 determines the target state based on the health status evaluation index values ​​acquired by the health status evaluation index value acquisition unit 192 for each state. The target setting unit 194 is an example of a target state setting means. For example, the target setting unit 194 may select as the target state the state in which the health status evaluation index value shows the best evaluation among the states within the search range.

[0067] The path search unit 195 searches for a path from the initial state to the target state. This path from the initial state to the target state can be considered a plan for improving health. The path search unit 195's search for a path can be seen as generating a plan. The route search unit 195 is an example of a route search means.

[0068] The path search unit 195 searches for a path from the initial state to the target state using evaluation values ​​for one or more state transitions, which are calculated using evaluation values ​​for one state transition. Here, as an evaluation for a single state transition, a better evaluation is given when the destination state is easier to achieve, and a better evaluation is given when the health status index value in the destination state is better than the health status index value in the source state.

[0069] The evaluation value for a single state transition indicates that the easier it is to achieve the destination state, which is expected to make it easier for the path search unit 195 to select a path that follows relatively easily achievable states. In other words, the path search unit 195 is expected to generate a plan that shows intermediate values ​​(intermediate target values) and target values ​​(final target values) for state identification items that are relatively easy for the subject to achieve.

[0070] Furthermore, the evaluation value for each state transition is expected to be better the health status evaluation index value in the destination state is compared to the health status evaluation index value in the source state, so that the path search unit 195 can generate a plan that allows the subject to check the results for each state transition.

[0071] It is expected that the ability for the subject to verify the results in the plan generated by the route search unit 195 will motivate the subject to continue executing the plan. With this motivation, it is expected that the subject will complete the plan and achieve the results anticipated in the plan.

[0072] The path search unit 195 may use an evaluation value for each state transition, which is a weighted sum of an evaluation value indicating that the destination state is easier to realize, and an evaluation value indicating that the health status evaluation index value in the destination state is better than the health status evaluation index value in the source state.

[0073] In this case, by adjusting the weights, the degree to which the path search unit 195 generates a plan that shows intermediate values ​​(intermediate target values) and target values ​​(final target values) for state identification items that are relatively easy for the subject to achieve, and the degree to which it generates a plan that allows the subject to check the results at each state transition, can be adjusted.

[0074] An evaluation value that indicates a better rating when the target state is easier to achieve is also called a feasibility evaluation value. An evaluation value that indicates a better rating when the health status evaluation value in the target state is better than the health status evaluation value in the source state is also called an outcome verifiability evaluation value.

[0075] The path search unit 195 may use the health status evaluation index value in the destination state as an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state. For transitions from the same source to each adjacent state to that source, whether an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state is used, or the health status evaluation index value in the destination state is used, the ranking of goodness of evaluation for each destination state will be the same.

[0076] After the target setting unit 194 sets the target state, the path search unit 195 may search for a path from the initial state to the target state. In this case, the path search unit 195 can terminate processing once it finds a path from the initial state to the target state, and does not need to search for paths to other states. In this respect, it is expected that the computational complexity of the path search unit 195 can be reduced relatively.

[0077] Alternatively, the path search unit 195 may search for paths from the initial state to each state within the search range, and then the target setting unit 194 may set the target state. In this case, the path search unit 195 can search for paths in advance (before the target setting unit 194 sets the target state). In this respect, the planning device 100 can generate a plan in real time from the setting of the target state.

[0078] Figure 2 shows an example of the display of feasibility evaluation values ​​by the display unit 120. Figure 2 shows an example where the condition identification items are weight and blood glucose, with the horizontal axis (horizontal direction in the figure) corresponding to the weight value and the vertical axis (vertical direction in the figure) corresponding to the blood glucose value. Furthermore, along the horizontal axis, the column numbers are indicated as 0, 1, 2, and 3 from left to right in the figure. A smaller column number (and therefore, further to the left in Figure 2) indicates a higher body weight. Along the vertical axis, the row numbers are indicated as 0, 1, and 2 from top to bottom. A smaller row number (and therefore, further to the top in Figure 2) indicates a higher blood glucose level. The display unit 120 may show a representative value or interval for body weight in each column and a representative value or interval for blood glucose in each row.

[0079] Furthermore, in the example in Figure 2, the probability of occurrence of each combination of weight and blood glucose value assigned to each state is shown as a feasibility assessment value. In this case, the larger the feasibility assessment index value, the more likely the subject is to achieve that state (the state indicated by that feasibility assessment index value). In the example in Figure 2, the initial state is the state at row 0 and column 0 (the leftmost and topmost state in Figure 2).

[0080] Figure 3 shows an example of the display of health status evaluation index values ​​by the display unit 120. Figure 3 shows an example where the condition identification items are weight and blood glucose, with the horizontal axis (horizontal direction in the figure) corresponding to the weight value and the vertical axis (vertical direction in the figure) corresponding to the blood glucose value. Furthermore, along the horizontal axis, the column numbers are indicated as 0, 1, 2, and 3 from left to right in the figure. A smaller column number (and therefore, further to the left in Figure 2) indicates a higher body weight. Along the vertical axis, the row numbers are indicated as 0, 1, and 2 from top to bottom. A smaller row number (and therefore, further to the top in Figure 2) indicates a higher blood glucose level. The display unit 120 may show a representative value or interval for body weight in each column and a representative value or interval for blood glucose in each row.

[0081] Furthermore, in the example in Figure 3, the health status evaluation index value is shown as 1 - risk of developing the disease (the value obtained by subtracting the risk of developing diabetes from 1). In this case, a higher health status evaluation index value indicates that the subject is in better health. In the example in Figure 3, the initial state is the state at row 0 and column 0 (the leftmost and topmost state in Figure 3).

[0082] Figure 4 shows an example of the display of evaluation index values ​​for each state by the display unit 120. Figure 4 shows an example where the condition identification items are weight and blood glucose, with the horizontal axis (horizontal direction in the figure) corresponding to the weight value and the vertical axis (vertical direction in the figure) corresponding to the blood glucose value. Furthermore, along the horizontal axis, the column numbers are indicated as 0, 1, 2, and 3 from left to right in the figure. A smaller column number (and therefore, further to the left in Figure 2) indicates a higher body weight. Along the vertical axis, the row numbers are indicated as 0, 1, and 2 from top to bottom. A smaller row number (and therefore, further to the top in Figure 2) indicates a higher blood glucose level. The display unit 120 may show a representative value or interval for body weight in each column and a representative value or interval for blood glucose in each row.

[0083] Furthermore, in the example in Figure 4, the evaluation index value for each condition is shown as probability of occurrence + 1 - risk of developing diabetes (the sum of the probability of occurrence of the weight and blood glucose combination assigned to each condition, and the risk of developing diabetes minus 1). In the example in Figure 4, the state in row 0 and column 0 (the leftmost and topmost state in Figure 4) is the initial state. The state in row 2 and column 3 (the rightmost and bottommost state in Figure 4) is the target state.

[0084] The path search unit 195 may use a value obtained by subtracting the evaluation index value for each state in the source state from the evaluation index value for each state in the destination state as the evaluation value for a single state transition. Furthermore, the path search unit 195 may select a path in which there are many transitions where the evaluation value for a single state transition is a positive value, and where the variation in the evaluation value for a single state transition between transitions is small. For example, the path search unit 195 may search for the path that maximizes the value of the evaluation function f shown in equation (8).

[0085]

number

[0086] x i x represents the evaluation value for the i-th state transition in the path (the evaluation value for a single state transition). i As an example, a value obtained by subtracting the evaluation index value for each state in the source state from the evaluation index value for each state in the destination state of the i-th state transition can be used. ReLU represents a Rectified Linear Unit, as shown in equation (9).

[0087]

number

[0088] The base of the logarithm (log) here is not limited to a specific value. For example, in equation (8), we could use the common logarithm or the natural logarithm. log(ReLU(x i )-2×ReLU(-x i )+5) is an example of an evaluation value for a single state transition. ReLU(x) in equation (8) i )-2×ReLU(-x i The value of ) is x i If ≥ 0, then x i And so, x i If <0, multiply by 2x i This is how it works. i If ≥ 0, ReLU(x i )-2×ReLU(-x i ) takes a positive value. x i If <0, ReLU(x i )-2×ReLU(-x i ) is ReLU(-x i Multiplying by a coefficient of -2 to x i It takes on a negative value that is larger in magnitude (absolute value) than when it is ≥ 0.

[0089] As a result, the decrease in the value of the evaluation function f in the case of a state transition that causes the health status evaluation index value to decrease is larger, and it is expected that the path search unit 195 will select a path with a relatively small number of state transitions that cause the health status evaluation index value to decrease. With a small number of state transitions that cause the health status evaluation index value to decrease in the plan, the subject can confirm more times during the execution of the plan that the health status evaluation index value has not decreased (i.e., that their health status has not deteriorated). As a result, it is expected that the subject's motivation to continue executing the plan will improve.

[0090] In equation (8), the "+5" is log(ReLU(x i )-2×ReLU(-x i The value of )+5) is made to be positive. log(ReLU(x i )-2×ReLU(-x iThe value of )+5) being positive allows the pathfinding unit 195 to use a pathfinding algorithm that requires the weights of the graph edges to be 0 or greater.

[0091] However, the "+5" in equation (8) is just one example, and the form of the equation and x i Depending on the possible values ​​of the variable, various values ​​can be added. Alternatively, in equation (8), a value greater than +5 could be added. In equation (8), both the feasibility assessment index value and the health assessment index value take values ​​within the range of 0 to 1, and -4 ≤ ReLU(x i )-2×ReLU(-x i Assuming that ) ≤ 2, "+5" is used. As in the example above, when the probability of occurrence is used as the feasibility assessment index value, the feasibility assessment index value takes a value within the range of 0 to 1. As in the example above, when the risk of onset is used as the health status assessment index value, the health status assessment index value takes a value within the range of 0 to 1. Alternatively, the pathfinding unit 195 may calculate evaluation index values ​​for each state before starting the pathfinding process, and then use ReLU(x i )-2×ReLU(-x i Alternatively, you could determine the value to add to ).

[0092] Also, in equation (8), (ReLU(x i )-2×ReLU(-x iBy taking the logarithm (log) of ()+5), it is expected that the increase in the health status evaluation index value will be suppressed (the increase in the evaluation function f will be smaller than when the logarithm is not taken) when the increase in the health status evaluation index value in a single state transition is large. As a result, it is expected that the path search unit 195 will detect paths in which the health status evaluation index value increases gradually with each state transition (rather than paths in which the health status evaluation index value increases significantly with a single state transition). As the health status evaluation index value increases gradually with each state transition, the subject can confirm that the health status evaluation index value is increasing (i.e., that their health is improving) during the execution of the plan. As a result, it is expected that the subject's motivation to continue executing the plan will improve.

[0093] Here, we consider a comparison between the case where the pathfinding unit 195 uses a logarithmic evaluation function f as shown in equation (8), and the case where it uses a non-logarithmic evaluation function f. If the pathfinding unit 195 uses an evaluation function f that does not take a logarithm, the pathfinding unit 195 may derive a path in which the evaluation function f takes a large value at a very small number of points along the path, and a small value at other points.

[0094] If the path search unit 195 derives such a path, there will be many cases where, during the execution of the plan, it will not be possible to confirm that the health status evaluation index value is increasing with each transition. As a result, the subject's motivation to continue executing the plan may not improve, or may even decrease.

[0095] In contrast, by using a logarithmic evaluation function f, as in the example of equation (8), the route search unit 195 is expected to increase the number of transitions in which the subject can confirm that the health status evaluation index value is increasing with each transition. This is expected to improve the subject's motivation to continue executing the plan.

[0096] In the example shown in Figure 4, the display unit 120 shows the path detected by the path search unit 195 by indicating each state transition with an arrow. However, the evaluation function used by the path search unit 195 for path searching is not limited to a specific one. In particular, the evaluation function used by the path search unit 195 for path searching is not limited to a logarithmic function. For example, the path search unit 195 may search for the path that maximizes the value of the evaluation function f shown in equation (10).

[0097]

number

[0098] ReLU(x i )-2×ReLU(-x i )+4 is an example of an evaluation value for a single state transition. The evaluation function f shown in equation (10) is an example of an evaluation function obtained by a method other than taking the logarithm. Specifically, the evaluation function f shown in equation (10) is an example obtained by removing the logarithmic operation from the evaluation function f shown in equation (8).

[0099] According to the evaluation function f shown in equation (10), the subexpression ReLU(x) of equation (8) is obtained. i )-2×ReLU(-x i Similar to the explanation given for the previous case, in the case of a state transition in which the health status evaluation index value decreases, the decrease in the value of the evaluation function f becomes larger, and it is expected that the path search unit 195 will select a path with a relatively small number of state transitions in which the health status evaluation index value decreases. With a small number of state transitions in the plan in which the health status evaluation index value decreases, the subject can confirm more times during the execution of the plan that the health status evaluation index value has not decreased (i.e., that their health status has not deteriorated). This is expected to improve the subject's motivation to continue executing the plan.

[0100] Note that the "+4" in equation (10) is just one example, and the form of the equation and x iDepending on the possible values ​​of the expression, various values ​​can be added. Alternatively, in equation (10), a value greater than +4 may be added. Furthermore, the path search unit 195 calculates evaluation index values ​​for each state before starting the path search, and then calculates ReLU(x i )-2×ReLU(-x i Alternatively, you could determine the value to add to ).

[0101] Alternatively, the path search unit 195 may search for the path that maximizes the value of the evaluation function f shown in equation (11).

[0102]

number

[0103] log(x i +3) is an example of an evaluation value for a single state transition. The evaluation function f shown in equation (11) is another example of an evaluation function using the logarithmic method. Specifically, the evaluation function f shown in equation (11) is an example obtained by removing the operation using ReLU from the evaluation function f shown in equation (8).

[0104] In this case, as explained with reference to equation (8), it is expected that when the increase in the health status evaluation index value in a single state transition is large, the increase in the value of the evaluation function f will be suppressed (the increase in the evaluation function f will be smaller than when the logarithm is not taken). As a result, it is expected that the path search unit 195 will detect a path in which the health status evaluation index value increases gradually with each state transition (rather than a path in which the health status evaluation index value increases significantly with a single state transition). As the health status evaluation index value increases gradually with each state transition, the subject can confirm that the health status evaluation index value is increasing (i.e., that their health is improving) during the execution of the plan. As a result, it is expected that the subject's motivation to continue executing the plan will improve.

[0105] Note that the "+3" in equation (11) is just one example, and the form of the equation and x i Depending on the possible values ​​of the expression, various values ​​can be added. Alternatively, in equation (11), a value greater than +3 may be added. Furthermore, the path search unit 195 calculates evaluation index values ​​for each state before starting the path search, and x i Alternatively, you could determine the value to add to it.

[0106] Furthermore, considering the case where the pathfinding unit 195 performs pathfinding using Dijkstra's algorithm, Dijkstra's algorithm requires that the weights of the graph edges be greater than or equal to 0, and that the weights represent costs. In other words, Dijkstra's algorithm uses weights with values ​​of 0 or greater to search for a path that minimizes the cumulative value of the weights. If the pathfinding unit 195 performs pathfinding using Dijkstra's algorithm and uses the natural logarithm as the logarithm, then equation (8) may be transformed into equation (12) so that the value of the evaluation function f represents the cost.

[0107]

number

[0108] -2≦x i If ≤ 2, then 1 ≤ ReLU(x i )-2×ReLU(-x i ) + 5 ≤ 7, and 2 - log(ReLU(x i )-2×ReLU(-x i ) + 5) ≥ 0. And x i The larger the value of (and therefore the better the evaluation), the smaller the value of the evaluation function f becomes. 2-log(ReLU(x i )-2×ReLU(-x i )+5) is an example of an evaluation value for a single state transition.

[0109] Figure 5 shows an example of a model representing health assessment index values ​​and feasibility assessment index values. That is, the health level evaluation index value acquisition unit 192 may use the model illustrated in FIG. 5 as a model for calculating the health level evaluation index value. Further, the feasibility evaluation index value acquisition unit 193 may use the model illustrated in FIG. 5 as a model for calculating the feasibility evaluation index value. FIG. 5 shows an example of expressing the health level evaluation index value and the feasibility evaluation index value using a hierarchical Bayesian model. N indicates the number of examinees. "N" in FIG. 5 indicates that the model is constructed using a training data set of N examinees.

[0110] y is a variable indicating the health level evaluation index value. For example, the flag Y indicating whether diabetes has developed within 3 years from the reference year, as described above n can be regarded as indicating the correct value of the variable y in the training data.

[0111] Also, in the description of the calculation of the diabetes onset risk described above, the training data for the year taken as the reference year is (X n , Y n )(n = 1,..., N), and an example of using this (X n , Y n ) when constructing a prediction model as a health level evaluation index was shown. And in the description of constructing a prediction model using a logistic regression model, X is used as an M-dimensional explanatory variable corresponding to the health examination data for the reference year, and Y is used as an objective variable representing whether diabetes has developed within 3 years from the reference year.

[0112] X cont Or X disc , or a combination thereof, corresponds to a variable for obtaining the feasibility evaluation index value. X cont represents an explanatory variable with a continuous value among the M-dimensional explanatory variables corresponding to the health examination data for the reference year (parameters taking values following a continuous distribution). Body weight and blood glucose correspond to examples of examination values following a continuous distribution. Xcont In relation to this, k indicates an index for identifying each continuous distribution that constitutes the mixture distribution. Here, a mixture Gaussian distribution is used as the continuous distribution. m k represents the mean value in each Gaussian distribution. Σ k represents the variance-covariance matrix in each Gaussian distribution.

[0113] The combination of α and θ represents the parameters of the mixture Gaussian distribution. α and θ correspond to examples of hidden variables. θ represents a K-dimensional vector, and the value of the k-th dimension is θ k ∈ {0, 1}. θ k = 1 is an indicator that points to the k-th mixture Gaussian distribution. Also, Σ k=1 K θ k = 1. Here, k is an integer such that 1 ≤ k ≤ K. α represents the mixing probability (the weight for each Gaussian distribution). Also, the probability that θ k = 1 is represented by α k That is, P(θ k = 1) = α k Here, 0 ≤ α k ≤ 1. P(θ k = 1) = α k Rewriting this with θ k as a probability can be expressed as in Equation (13).

[0114]

Equation

[0115] The conditional probability of X by the K-th Gaussian distribution can be expressed as in Equation (14).

[0116]

Equation

[0117] Using the vector θ, it can be expressed as in Equation (15).

[0118]

number

[0119] Furthermore, the marginal distribution P(X) of X can be expressed using P(X|θ)P(θ) as shown in equation (16).

[0120]

number

[0121] P(X) can be expressed as shown in equation (17).

[0122]

number

[0123] X disc This indicates parameters that take discrete values ​​(parameters that follow a discrete distribution) among the parameters used in calculating health status evaluation index values ​​and feasibility evaluation index values. Gender and responses to questionnaire items conducted during health checkups are examples of discrete distributions. X disc In this relationship, k represents an index that identifies the individual discrete distributions that make up the mixture distribution. Here, categorical distributions are used as the discrete distributions. φ k φ represents the parameter of the k-th categorical distribution. The parameters of the K categorical distributions are given by φ=(φ1,…,φ k ,…,φ K ) is expressed as. Here, if X is the data of the patients who underwent health checkups, which are M-dimensional explanatory variables corresponding to the health checkup data of the base year and have continuous explanatory variables, then P(X) represents the probability of X occurring. This probability of occurrence P(X) is expressed as the probability P(X^1=X^) in equation (6). i 1, X^2=X^ jIt may be treated similarly to (2), and P(X) may be used as a feasibility assessment index value.

[0124] The combination of τ and γ represents the parameters of a mixture of categorical distributions. τ and γ are examples of latent variables. The combination of τ and γ is similar to the combination of the parameters α and θ in a mixture of Gaussian distributions. γ represents a K-dimensional vector, and the value of the k-th dimension is γ k ={0,1}. γ k =1 is an indicator that points to the k-th Gaussian mixture distribution. Also, Σ k=1 K γ k = 1. τ represents the mixed probability (the weight of each category on the Gaussian distribution).

[0125] As described above, using logistic regression, the conditional probability P(Y=1|X;W) for Y=1 is given by equation (1) above. This conditional probability P may be treated in the same way as the probability P(Y=1|X;W*) described above, or the conditional probability P may be used as a health assessment index value. W T X is obtained by summing the products of the elements of the weight vector W and the elements of the explanatory variable vector X over M dimensions, and is shown in equation (18).

[0126]

number

[0127] In Figure 5, the weight vector W is represented by β. Figure 5 also shows an example where M=3. Furthermore, Figure 5 can be seen as representing P(Y=1|X;W) as a mixture logistic regression model. k is an index that identifies the individual logistic regressions that make up the mixture distribution. β kThis represents the parameters of the k-th logistic regression. In Figure 5, β k =( β 1,k ,β 2,k ,β 3,k ) The combination of π and z represents the parameters of a logistic regression mixture. This combination of π and z is similar to the combination of α and θ for the parameters of a Gaussian mixture.

[0128] Figure 6 shows an example of the process by which the planning device 100 generates and outputs a plan. In the process shown in Figure 6, the data acquisition unit 191 acquires various data for generating the plan (step S101). Next, the processing unit 190 sets the search range (step S102).

[0129] Next, the health status evaluation index value acquisition unit 192 calculates the health status evaluation index value for each state within the search range (step S103). Furthermore, the feasibility evaluation index value acquisition unit 193 calculates the feasibility evaluation index value for each state within the search range (step S104).

[0130] Next, the path search unit 195 searches for a path from the initial state to each state within the search range using an evaluation function that uses a health evaluation index value for each state and a feasibility evaluation index value for each state (step S105). The path search method used by the path search unit 195 is not limited to a specific one. For example, the path search unit 195 may use Dijkstra's algorithm to perform the path search, but is not limited to this.

[0131] Next, the target setting unit 194 sets the target state (step S106). The target setting unit 194 may determine the target state to be the state in which the health status evaluation index value indicates the best evaluation among the states within the search range (i.e., the state in which the health status evaluation index value indicates the best health status).

[0132] Next, the planning device 100 outputs a path from the initial state to the target state as a plan (step S107). For example, the path search unit 195 may select a path to the target state from among the paths detected for each state within the search range and treat it as a plan. Then, the processing unit 190 may control the display unit 120 to display the plan. After step S107, the planning device 100 completes the process shown in Figure 6.

[0133] Figure 7 shows an example of the procedure for the pathfinding unit 195 to perform a path search. Figure 7 shows an example where the pathfinding unit 195 performs a path search using Dijkstra's algorithm. When the pathfinding unit 195 uses Dijkstra's algorithm, it uses a cost that takes a positive value as the evaluation value for a single state transition, as in the example of equation (12). The route search unit 195 performs the process shown in Figure 7 in step S105 of Figure 6. In the process shown in Figure 7, the route search unit 195 performs initial setup for route searching (step S201).

[0134] The route search unit 195 sets the initial values ​​of the list PL, the route evaluation index eval[v] for each state v, and the previous state pred[v] for each state v during the initial setup. The path evaluation index eval[v] is a variable that indicates the evaluation value of the path from the initial state s to state v.

[0135] The list PL is a list whose elements represent the state where the value of the path evaluation index eval[v] is undetermined. The previous state pred[v] is a variable that indicates the state immediately preceding state v (the state that transitions directly to state v) in the path from the initial state s to state v.

[0136] Figure 8 shows an example of the procedure for the route search unit 195 to perform initial settings for route searching. The route search unit 195 performs the process shown in Figure 8 in step S201 of Figure 7. In the process shown in Figure 8, the route search unit 195 makes the list PL an empty list (step S211).

[0137] Next, the pathfinding unit 195 starts a loop L11 that processes each state v included in the search range V (step S212). The state being processed in loop L11 is denoted as state v. In the processing in loop L11, the path search unit 195 sets the value of the path evaluation index eval[v] to infinity (∞) (step S213). The processing in step S203 can be expressed as eval[v]:=∞. Note that the value that the path search unit 195 sets for the path evaluation index eval[v] in step S213 is not limited to infinity, but can be any value that is sufficiently large compared to the actual calculated value of the path evaluation index eval[v].

[0138] Furthermore, the pathfinding unit 195 sets the value of the previous state pred[v] to -1 (step S214). The process in step S204 can be expressed as pred[v]:=-1. A value of -1 for the previous state pred[v] indicates that the value of the previous state pred[v] is undetermined. The case where the value of the previous state pred[v] is undetermined includes the case where there is no previous state for state v.

[0139] Furthermore, the pathfinding unit 195 inserts state v into list PL (step S215). That is, the pathfinding unit 195 includes state v as an element of list PL.

[0140] Next, the pathfinding unit 195 performs termination processing for loop L11 (step S216). Specifically, the pathfinding unit 195 determines whether or not it has performed processing for all states v included in the search range V. If it determines that there are states v for which loop L11 processing has not been performed, the pathfinding unit 195 continues to perform loop L11 processing for the unprocessed states v. On the other hand, if it determines that it has performed processing for all states v included in the search range V, the pathfinding unit 195 terminates loop L11.

[0141] If the loop L11 is terminated in step S216, the pathfinding unit 195 sets the value of the path evaluation index eval[s] in the initial state s to 0 (step S217). The process in step S217 can be expressed as eval[s]:=0. In steps S213 and S217, the path search unit 195 initializes the value of the path evaluation index eval[s] for the initial state s to 0, and sets the value of the path evaluation index eval[v] for all other states v to infinity.

[0142] Furthermore, the path search unit 195 calculates an evaluation value for each combination of two adjacent states within the search range (step S218). If the two adjacent states are designated as the first state and the second state, and both a state transition from the first state to the second state and a state transition from the second state to the first state are possible, the path search unit 195 calculates an evaluation value for each of these two state transitions.

[0143] The evaluation value for a single state transition corresponds to the weight of the edges in the graph. In Dijkstra's algorithm, the path search unit 195 calculates the value of the path evaluation index eval[v] by accumulating the evaluation values ​​for each state transition. After step S218, the route search unit 195 completes the process shown in Figure 8. In this case, the route search unit 195 completes the process shown in step S201 in Figure 7, and the process proceeds to step S221.

[0144] After step S201 in Figure 7, the path search unit 195 selects the state v in the list PL that has the minimum value of the path evaluation index eval[v], and sets it to state u (step S221). The process in step S221 can be expressed as shown in equation (19).

[0145]

number

[0146] The state u can be regarded as the state reached at the current point in the path search. The state u is also referred to as the current state in the path search. Next, the path search unit 195 excludes the state u from the list PL (step S222). That is, the path search unit 195 deletes the state u among the elements of the list PL.

[0147] Next, the path search unit 195 determines whether the list PL is empty (step S223). If it is determined that the list PL is not empty (step S223: NO), the path search unit 195 starts a loop L12 for performing processing for each state v adjacent to the state u (step S231). The state being processed in the loop L12 is denoted as the state v.

[0148] In the processing in the loop L12, the path search unit 195 calculates the path evaluation index value when transitioning from the state u to the state v (step S232). The path evaluation index value when transitioning from the state u to the state v is also denoted as newEval. The path search unit 195 calculates the path evaluation index value newEval when transitioning from the state u to the state v by adding the evaluation value for one state transition from the state u to the state v to the value of the path evaluation index eval[u] in the state u. The evaluation value for one state transition from the state u to the state v is denoted as x i Then, the processing in step S232 can be expressed as in Equation (20).

[0149]

Equation

[0150] As described above, as the evaluation value x for one state transition i a cost that takes a positive value is used. Next, the path search unit 195 determines whether newEval < eval[v] (step S233). When it is determined that newEval < eval[v], the path search unit 195 updates the value of the path evaluation index eval[v] in state v to the path evaluation index value newEval when transitioning from state u to state v (step S234). The process in step S234 can be expressed as eval[v] := newEval.

[0151] Also, the path search unit 195 updates the previous state pred[v] of state v to state u (step S235). The process in step S235 can be expressed as pred[v] := u.

[0152] Next, the path search unit 195 performs the end process of loop L12 (step S236). Specifically, the path search unit 195 determines whether the process of loop L12 has been performed for all states v adjacent to the current state u in the path search. If it is determined that there is a state v for which the process of loop L12 has not been performed, the path search unit 195 continues to perform the process of loop L12 for the unprocessed state v. On the other hand, if it is determined that the process of loop L12 has been performed for all states v adjacent to the current state u in the path search, the path search unit 195 ends loop L12. When the path search unit 195 ends loop L12 in step S236, the process returns to step S221.

[0153] On the other hand, when the path search unit 195 determines in step S233 that newEval ≥ eval[v] (step S233: NO), the process proceeds to step S236. On the other hand, when it is determined in step S223 that the list PL is empty (step S223: YES), the path search unit 195 ends the process of FIG. 7. In this case, the path search unit 195 ends the process of step S105 in FIG. 6, and the process proceeds to step S106.

[0154] The path search unit 195 may perform a path search after the goal setting unit 194 has set the goal state. In this case, the feasibility evaluation index value acquisition unit 193 may acquire the feasibility evaluation index value when necessary during the path search. Figure 9 shows an example of the procedure for the planning device 100 to generate and output a plan when the target setting unit 194 sets the target state and the path search unit 195 performs a path search.

[0155] Steps S301 to S303 in Figure 9 are the same as steps S101 to S103 in Figure 6. After step S303, the target setting unit 194 sets the target state (step S304). Step S304 is the same as step S106 in Figure 6.

[0156] The process from steps S305 to S309 differs from the processes in steps S104 to S106 in Figure 6, Figures 7 and 8 in that, during the path search, the feasibility evaluation index value acquisition unit 193 calculates an evaluation value for state transitions, the path search unit 195 calculates an evaluation value for a single state transition, and the path search unit 195 terminates the path search when it reaches the target state. In all other respects, the process from steps S305 to S309 is the same as the process in steps S104 to S106 in Figure 6, Figures 7 and 8.

[0157] After step S304, the route search unit 195 performs initial settings for route search (step S305). Step S305 is the same as steps S211 to S218 in Figure 8. Next, the pathfinding unit 195 selects state u from the states included in the list PL (step S306). Step S306 is the same as step S221 in Figure 7. Next, the pathfinding unit 195 removes state u from list PL (step S307). Step S307 is the same as step S222 in Figure 7.

[0158] Next, the path search unit 195 determines whether or not the target state has been reached (step S308). Specifically, the path search unit 195 determines whether or not the state u matches the target state. If the path search unit 195 determines that the target state has not been reached (step S308: NO), the feasibility evaluation index value acquisition unit 193 acquires the feasibility evaluation index values for each state adjacent to the state u (the current state in the path search) among the states included in the list PL (step S311). If there is a state for which the feasibility evaluation index value has already been acquired, the already acquired feasibility evaluation index value can be used, and it is not necessary for the feasibility evaluation index value acquisition unit 193 to acquire the feasibility evaluation index value for that state again.

[0159] Next, for each state adjacent to the state u, the path search unit 195 calculates the evaluation value x i for a single state transition from the state u (step S312). Next, the path search unit 195 executes a path search algorithm for the state u selected in step S306 (step S313). Step S313 is the same as steps S231 to S236 in FIG. 7. The feasibility evaluation index value acquisition unit 193 and the path search unit 195 may perform the processing of steps S311 and S312 within a loop corresponding to loop L12 in FIG. 7. After step S313, the process returns to step S306. On the other hand, if in step S308, the path search unit 195 determines that the target state has been reached (step S308: YES), the planning device 100 outputs a plan (step S321). Step S321 is the same as step S107 in FIG. 6. After step S321, the planning device 100 ends the process of FIG. 9.

[0160] The planning device 100 may expand the search range. For example, when the subject determines the target value of the health degree evaluation index, if there is no state within the search range that satisfies the target value, the planning device 100 may expand the search range. FIG. 10 is a diagram showing an example of the procedure of the planning device 100 generating and outputting a plan when expanding the search range.

[0161] Step S401 in FIG. 10 is the same as step S101 in FIG. 6. After step S401, the data acquisition unit 191 determines the target value of the health level evaluation index (step S402). For example, the data acquisition unit 191 may set the target value received by the operation input unit 130 through a user operation as the target value in the route search.

[0162] Steps S403 to S404 are the same as steps S102 to S103 in FIG. 6. Next, the target setting unit 194 determines whether there is a state within the search range that satisfies the target value set in step S402 (step S405).

[0163] If the target setting unit 194 determines that there is no state that satisfies the target value (step S405: NO), the processing unit 190 expands the search range (step S411). For example, the processing unit 190 may widen the range of each state identification item value as the search range by a predetermined width. By expanding the search range, the number of states included in the search range increases.

[0164] Next, the health level evaluation index value acquisition unit 192 acquires the tendency degree evaluation index values for each state that has newly become included in the search range due to the expansion of the search range (step S412). After step S412, the process returns to step S405.

[0165] On the other hand, if in step S405, it is determined that there is a state within the search range that satisfies the target value set in step S402 (step S405: YES), the target setting unit 194 sets the state determined to satisfy the target value as the target state (step S421). If there are multiple states that satisfy the target value, the target setting unit 194 sets any one of those multiple states as the target state.

[0166] Next, the path search unit 195 searches for a path from the initial state to the target state (step S422). Step S422 is the same as steps S305 to S313 in Figure 9. Similar to the process in Figure 6, the health status evaluation index acquisition unit 192 and the path search unit 195 may calculate the health status evaluation index value and the evaluation value for a single state transition before executing the path search loop. For example, the feasibility evaluation index acquisition unit 193 may, in step S412, acquire the feasibility evaluation index value for each state that has newly been included in the search range due to the expansion of the search range, instead of the process corresponding to step S311 in Figure 9. Alternatively, the path search unit 195 may, in step S412, calculate the evaluation value for a single state transition for each state transition from each state that has newly been included in the search range due to the expansion of the search range, and for each state transition to each of these states, instead of the process corresponding to step S312 in Figure 9.

[0167] After step S422, the planning device 100 outputs the plan (step S423). The transition from step S422 to S423 corresponds to the transition to step S321 in step S308 of Figure 9 if the answer is YES. Step S423 is the same as step S321 in Figure 9. After step S423, the planning device 100 completes the process shown in Figure 10.

[0168] The planning device 100 may generate plans that span multiple periods. These periods are not limited to any particular one. For example, the planning device 100 may generate plans that span multiple years, with each year considered as one period. Figure 11 shows an example of the procedure for the planning device 100 to generate and output a plan that spans multiple periods.

[0169] In the process shown in Figure 11, the data acquisition unit 191 acquires various data for generating the plan (step S501). In particular, the data acquisition unit 191 acquires data for multiple periods. For example, the data that should be referenced for plan generation may differ depending on the period, such as as the age of the subject increases. By the data acquisition unit 191 acquiring data for multiple periods, it is expected that the planning device 100 will be able to generate the plan with higher accuracy. Step S501 is the same as step S101 in Figure 6, except that the data acquisition unit 191 acquires data for multiple periods.

[0170] Next, the processing unit 190 sets the search range (step S502). Step S502 is the same as step S102 in Figure 6. Next, the health status evaluation index value acquisition unit 192 acquires the health status evaluation index value for each state within the search range for each period (step S503).

[0171] By assigning states to each period, the transition between periods can be represented as state transitions, and processing can be performed in the same way as when the planned period is treated as a single period, as shown in the example in Figure 6. However, a constraint is imposed on the state transitions due to the change of periods: it is not possible to transition from a state in the new period to a state in the old period. Furthermore, by having the health status evaluation index value acquisition unit 192 acquire health status evaluation index values ​​for each state within the search range at each time period, it is expected that the planning device 100 will be able to generate plans with higher accuracy.

[0172] Next, the feasibility evaluation index value acquisition unit 193 acquires feasibility evaluation index values ​​for each state within the search range for each period (step S504). By the feasibility evaluation index value acquisition unit 193 acquiring feasibility evaluation index values ​​for each state within the search range for each period, the planning device 100 is expected to be able to generate plans with higher accuracy.

[0173] Steps S505 to S507 are the same as steps S105 to S107 in FIG. 6. When ensuring that the plan always covers a predetermined period, the target setting unit 194 selects the target state from the last period among the plurality of periods targeted by the plan. After step S507, the planning device 100 ends the process of FIG. 11.

[0174] FIG. 12 is a diagram showing an example of a model of the feasibility evaluation index used by the feasibility evaluation index value acquisition unit 193 when the planning device 100 generates a plan spanning a plurality of periods. FIG. 12 shows an example when the planning device 100 generates a plan spanning two periods, where t represents the older of the two periods and t + 1 represents the newer of the two periods.

[0175] In the example of FIG. 12, the model in the example of FIG. 5 is provided for each period. And for the parameter X that takes continuous values in period t cont t and the parameter X that takes continuous values in period t + 1 cont t+1 are assumed to have a relationship (not independent). Also, for the parameter X that takes discrete values in period t disc t and the parameter X that takes discrete values in period t + 1 disc t+1 are assumed to have a relationship (not independent). Parameter X cont t and parameter X cont t+1 The relationship between them can be expressed as in Equation (21).

[0176]

Equation

[0177] Equation (21) is for the parameter X that takes continuous values in period t cont tGiven this, what is the continuous parameter X over the period t+1? cont t+1 This shows that the conditional distribution of follows a Gaussian distribution. Here, AX cont t σI represents the expected value of the Gaussian distribution, and σI represents the variance-covariance matrix of the Gaussian distribution. Here, A represents an M x 1-dimensional matrix, and σI represents the M-dimensional identity matrix.

[0178] Parameter X disc t and parameter X disc t+1 The relationship between them can be expressed by the fact that the transition probability is given by equation (22).

[0179]

number

[0180] Figure 13 shows an example of displaying a plan spanning multiple periods using the display unit 120. Figure 13 shows an example where the condition identification items are weight and blood glucose, and the period is a year. In the graph in Figure 13, the x-axis represents blood glucose, the y-axis represents weight, and the z-axis represents the year. In the example in Figure 13, both weight and blood glucose are shown as the difference from the reference value. Point P111 indicates the initial state. Lines L111, L112, and L113 all represent the routes in the first year. Point P112 represents the target state for the first year. Point P112 can be considered an intermediate target state in the plan.

[0181] Line L121 represents the state transition corresponding to the shift from the first year to the second year. Lines L122 and L123 both indicate the routes for the second year. Point P121 represents the target state for the second year. Point P121 can be considered an intermediate target state in the plan.

[0182] Line L131 represents the state transition corresponding to the shift from the second year to the third year. Line L132 shows the route in the third year. Point P131 indicates the target state.

[0183] Figure 14 shows a first example of the display of period-specific data by the display unit 120. Figure 14 shows data for the first period (the older period) of the two periods being planned. Figure 14 shows an example where the condition identification items are weight and blood glucose, with the horizontal axis (horizontal direction in the figure) corresponding to the weight value and the vertical axis (vertical direction in the figure) corresponding to the blood glucose value. Furthermore, along the horizontal axis, the column numbers are indicated as 0, 1, 2, and 3 from left to right as you view the figure. A smaller column number (and therefore, further to the left as you view Figure 14) indicates a higher body weight. Along the vertical axis, the row numbers are indicated as 0, 1, and 2 from top to bottom. A smaller row number (and therefore, further to the top of Figure 14) indicates a higher blood glucose level. The display unit 120 may show a representative value or interval for body weight in each column and a representative value or interval for blood glucose in each row.

[0184] Furthermore, in the example in Figure 14, the evaluation index value for each condition is shown as probability of occurrence + 1 - risk of developing diabetes (the sum of the probability of occurrence of the weight and blood glucose combination assigned to each condition, and the risk of developing diabetes minus 1). In the example in Figure 14, the initial state is the state at row 0 and column 0 (the leftmost and topmost state in Figure 14).

[0185] Figure 15 shows a second example of the display of period-specific data by the display unit 120. Figure 15 shows data for the second period (the more recent period) of the two periods being planned. Figure 15 shows an example where the condition identification items are weight and blood glucose, with the horizontal axis (horizontal direction in the figure) corresponding to the weight value and the vertical axis (vertical direction in the figure) corresponding to the blood glucose value. Furthermore, along the horizontal axis, the column numbers are indicated as 0, 1, 2, and 3 from left to right as you view the figure. A smaller column number (and therefore, further to the left in Figure 15) indicates a higher body weight. Along the vertical axis, the row numbers are indicated as 0, 1, and 2 from top to bottom. A smaller row number (and therefore, further to the top in Figure 15) indicates a higher blood glucose level. The display unit 120 may show a representative value or interval for body weight in each column and a representative value or interval for blood glucose in each row.

[0186] Furthermore, in the example in Figure 15, the evaluation index value for each condition is shown as probability of occurrence + 1 - risk of developing diabetes (the sum of the probability of occurrence of the weight and blood glucose combination assigned to each condition, and the risk of developing diabetes minus 1). Figures 14 and 15 show data for the same range of weight and blood glucose values ​​for the first and second periods, respectively. In this way, it is expected that the planning device 100 can generate plans with relatively high accuracy by acquiring data for each period for the same state-specific item values.

[0187] As described above, the feasibility evaluation index value acquisition unit 193 acquires feasibility evaluation index values ​​that quantitatively indicate the ease of achieving each individual state, for each state identified using the values ​​of one or more items that correlate with the health status evaluation index value. The health status evaluation index value acquisition unit 192 acquires health status evaluation index values ​​for each individual condition. The path search unit 195 uses an evaluation value for a path consisting of one or more state transitions, calculated using evaluation values ​​for each individual state transition, where the evaluation for a single state transition from the source state to the destination state is better if the destination state is easier to realize, and the evaluation is better if the health status evaluation index value in the destination state is better than the health status evaluation index value in the source state. The path search unit searches for a path from the initial state to the target state in the plan using this evaluation value for a path consisting of one or more state transitions.

[0188] According to the planning device 100, it is expected that the person executing the plan, which involves gradually changing a combination of quantitative values ​​designated as state-specific items, will be able to confirm the results of the plan's execution. In particular, the planning device 100 uses evaluation values ​​for each state transition, where the health status evaluation index value in the destination state is better than the health status evaluation index value in the source state, thus allowing the subject to confirm the change in the health status evaluation index value during state transitions. Furthermore, the planning device 100 uses evaluation values ​​that indicate a better evaluation the easier it is to achieve the target state, which is expected to make it relatively easy for the user to execute the plan.

[0189] Furthermore, the path search unit 195 uses the health status evaluation index value in the destination state as an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state. According to the planning device 100, it is expected that a state transition that improves the evaluation indicated by the health status evaluation index value can be selected with relatively simple calculations.

[0190] Furthermore, the goal setting unit 194 determines the target state in the plan based on the health status evaluation index value for each state. The path search unit 195 searches for a path from the initial state in the plan to the target state determined by the target setting unit 194. According to the planning device 100, the path search can be terminated once the target state is reached. In this respect, the planning device 100 is expected to require relatively little computation.

[0191] Furthermore, the path search unit 195 searches for paths from the initial state in the plan to individual states other than the initial state. Then, among the paths detected by the path search unit 195, the path from the initial state to the state set as the target state is used as the path from the initial state to the target state in the plan. According to the planning device 100, pathfinding can be performed before the target state is set. In this respect, the planning device 100 is expected to shorten the time from setting the target state to generating the plan.

[0192] Furthermore, the health status evaluation index value acquisition unit 192 acquires health status evaluation index values ​​for each state that can be reached within a predetermined number of state transitions from the initial state in the plan. According to the planning device 100, the search range can be specified by specifying the number of state transitions from the initial state. In this respect, the planning device 100 makes it relatively easy to specify the search range.

[0193] Furthermore, the path search unit 195 uses an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each state transition, which is weighted and summed by an evaluation value for each state transition. This evaluation value is calculated by adding together an evaluation value for each state transition, which indicates a better evaluation the easier it is to realize the destination state, and an evaluation value for which the health evaluation index value in the destination state is better than the health evaluation index value in the source state. The path search unit 195 then searches for a path from the initial state to the target state in the plan. According to the planning device 100, the weighting of the ease of executing the plan and the likelihood that the subject will be able to confirm the results of the plan's execution can be adjusted through a relatively simple process of adjusting the weights.

[0194] Furthermore, the health status evaluation index value acquisition unit 192 calculates the health status evaluation index value using a model in which the parameters of the health status evaluation index, specifically the parameters that take values ​​following a continuous distribution and the parameters that take values ​​following a discrete distribution, have different hidden variables as parameters. According to the planning device 100, it is expected that feasibility evaluation index values ​​can be calculated with relatively high accuracy.

[0195] Furthermore, the status in the plan is identified using the values ​​of one or more items that correlate with the health status assessment index value, and one of several time periods. The feasibility assessment index value acquisition unit 193 acquires feasibility assessment index values ​​for each value of one or more items that correlate with the health status assessment index value, and for each period. The health status evaluation index value acquisition unit 192 acquires the health status evaluation index value for each value of one or more items that correlate with the health status evaluation index value, and for each period. According to the planning device 100, it is expected that plans can be generated with relatively high accuracy by acquiring health status evaluation index values ​​and feasibility evaluation index values ​​for each period.

[0196] Furthermore, the health status evaluation index value acquisition unit 192 receives input values ​​for one or more items that identify a state and calculates the health status evaluation index value for each individual state using a trained model that outputs the health status evaluation index value for the identified state. According to the planning device 100, by acquiring a model that outputs health status evaluation index values ​​through learning, it is possible to obtain health status evaluation index values ​​that reflect statistical data. In this respect, the planning device 100 is expected to be able to generate plans with relatively high accuracy.

[0197] Furthermore, one or more items used to identify the condition of the target individuals in the plan are selected based on the correlation between each measured item and the health status evaluation index value. According to the planning device 100, since the items used to identify the state (state identification items) are selected based on the correlation between the measured items and the health status evaluation index values, it is expected that changes in the state identification item values ​​will affect the health status evaluation index values. In this respect, it is expected that the planning device 100 can generate plans with relatively high accuracy.

[0198] <Second Embodiment> Figure 16 shows an example of the configuration of a planning device according to at least one embodiment. In the configuration shown in Figure 16, the planning device 610 includes a feasibility evaluation index value acquisition unit 611, a health status evaluation index value acquisition unit 612, and a route search unit 613.

[0199] In this configuration, the feasibility evaluation index value acquisition unit 611 acquires a feasibility evaluation index value that quantitatively indicates the ease of achieving each individual state, for each state identified using the values ​​of one or more items that correlate with the health status evaluation index value. The health status evaluation index value acquisition unit 612 acquires health status evaluation index values ​​for each individual condition. The path search unit 613 searches for a path from the initial state to the target state in the plan using an evaluation value for a path consisting of one or more state transitions, which is calculated using an evaluation value for each state transition, where the evaluation is given more favorably the easier it is to realize the target state, and the evaluation is given more favorably the health status evaluation index value in the target state is better than the health status evaluation index value in the source state. The feasibility evaluation index value acquisition unit 611 is an example of a means for acquiring feasibility evaluation index values. The health status evaluation index value acquisition unit 612 is an example of a means for acquiring health status evaluation index values. The route search unit 613 is an example of a means for searching for a route.

[0200] According to the planning device 610, it is expected that the person executing the plan, which involves gradually changing a combination of quantitative values ​​designated as state identification items (items used to identify a state), will be able to confirm the results of the plan execution. In particular, the planning device 610 uses evaluation values ​​for each state transition, indicating that a better evaluation is given when the health status evaluation index value in the destination state is better than the health status evaluation index value in the source state. This allows the subject to be able to confirm the changes in the health status evaluation index value during state transitions. Furthermore, the planning device 610 uses evaluation values ​​that indicate a better evaluation the easier it is to achieve the transition state, which is expected to make it relatively easy for the user to execute the plan.

[0201] <Third Embodiment> Figure 17 shows an example of the processing steps in a planning method according to at least one embodiment. The planning method shown in Figure 17 includes obtaining feasibility assessment index values ​​(step S611), obtaining health assessment index values ​​(step S612), and searching for a route (step S613).

[0202] In obtaining feasibility assessment index values ​​(step S611), the computer obtains feasibility assessment index values ​​that quantitatively indicate the ease of achieving each individual state, for each state identified using the values ​​of one or more items that correlate with the health status assessment index values. In obtaining health status evaluation index values ​​(step S612), the computer obtains health status evaluation index values ​​for each individual condition. In the process of searching for a path (step S613), the computer searches for a path from the initial state to the target state in the plan using evaluation values ​​for one or more state transitions, which are calculated using evaluation values ​​for each state transition, where the evaluation is given more favorably the easier it is to realize the target state, and the evaluation is given more favorably the health status evaluation index value in the target state is better than the health status evaluation index value in the source state.

[0203] According to the planning method shown in Figure 17, it is expected that the person executing the plan, which involves gradually changing combinations of quantitative values ​​designated as state-specific items (items used to identify a state), will be able to confirm the results of the plan's execution. In particular, according to the planning method shown in Figure 17, an evaluation value is used for each state transition, where a better evaluation is given when the health status evaluation index value in the destination state is compared to the health status evaluation index value in the source state. This allows the subject to be able to confirm the change in the health status evaluation index value during state transitions. Furthermore, according to the planning method shown in Figure 17, since it uses evaluation values ​​that indicate a better evaluation the easier it is to achieve the target state, it is expected that the target users will find it relatively easy to implement the plan.

[0204] Figure 18 shows an example of a computer configuration according to at least one embodiment. In the configuration shown in Figure 18, the computer 700 comprises a CPU 710, a main memory 720, an auxiliary memory 730, an interface 740, and a non-volatile recording medium 750.

[0205] One or more of the above-described planning devices 100 and 610, or parts thereof, may be implemented in the computer 700. In that case, the operation of each processing unit described above is stored in the auxiliary storage device 730 in the form of a program. The CPU 710 reads the program from the auxiliary storage device 730, expands it in the main memory 720, and executes the above-described processing according to the program. The CPU 710 also reserves memory areas in the main memory 720 corresponding to each of the above-described storage units according to the program. Communication between each device and other devices is performed by the interface 740 having a communication function and communicating according to the control of the CPU 710. The interface 740 also has a port for the non-volatile recording medium 750 and reads information from the non-volatile recording medium 750 and writes information to the non-volatile recording medium 750.

[0206] When the planning device 100 is implemented in the computer 700, the operation of the processing unit 190 and each of its parts is stored in auxiliary storage device 730 in the form of a program. The CPU 710 reads the program from the auxiliary storage device 730, loads it into the main memory 720, and executes the above processing according to the program.

[0207] Furthermore, the CPU 710 reserves a memory area for the memory unit 180 in the main memory 720 according to the program. Communication with other devices by the communication unit 110 is performed by the interface 740 having a communication function and operating under the control of the CPU 710. Display of images by the display unit 120 is performed by the interface 740 having a display device and displaying various images under the control of the CPU 710. Acceptance of user operations by the operation input unit 130 is performed by the interface 740 having an input device and accepting user operations under the control of the CPU 710.

[0208] When the planning device 610 is implemented in the computer 700, the operations of the feasibility evaluation index value acquisition unit 611, the health status evaluation index value acquisition unit 612, and the path search unit 613 are stored in auxiliary storage device 730 in the form of a program. The CPU 710 reads the program from the auxiliary storage device 730, loads it into the main memory device 720, and executes the above processes according to the program.

[0209] Furthermore, the CPU 710 reserves memory in the main memory 720 for the planning device 610 to process, according to the program. Communication between the planning device 610 and other devices is performed by the interface 740 having a communication function and operating under the control of the CPU 710. Interaction between the planning device 610 and the user is performed by the interface 740 having input and output devices, presenting information to the user via the output device and accepting user operations via the input device under the control of the CPU 710.

[0210] One or more of the above-mentioned programs may be recorded on the non-volatile recording medium 750. In this case, the interface 740 may read the program from the non-volatile recording medium 750. The CPU 710 may then either directly execute the program read by the interface 740, or temporarily save it in the main memory 720 or auxiliary memory 730 before executing it.

[0211] Alternatively, a program for executing all or part of the processing performed by the planning device 100 and the planning device 610 may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be loaded into a computer system and executed to perform the processing of each part. The term "computer system" here includes hardware such as the OS (Operating System) and peripheral devices. Furthermore, "computer-readable recording media" refers to portable media such as flexible disks, magneto-optical disks, ROMs (Read Only Memory), CD-ROMs (Compact Disc Read Only Memory), and storage devices such as hard disks built into computer systems. The above-mentioned program may be intended to implement only a part of the functions described above, and may also be able to implement the above-mentioned functions in combination with programs already recorded in the computer system.

[0212] Although the present disclosure has been described above with reference to these embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure are possible, as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, the embodiments described above may be combined with other embodiments as appropriate.

[0213] Some or all of the above embodiments may also be described as follows, but are not limited to these.

[0214] (Note 1) A feasibility evaluation index value acquisition means for acquiring a feasibility evaluation index value that quantitatively indicates the ease of achieving each of the states identified using the values ​​of one or more items that correlate with the health status evaluation index value, A means for acquiring health status evaluation index values ​​for acquiring the health status evaluation index values ​​for each of the aforementioned states, A path search means for searching a path from an initial state to a target state in a plan, using an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each individual state transition, wherein the evaluation for a single state transition from the source state to the destination state is such that the easier it is to realize the destination state, the better the evaluation, and the better the health status evaluation index value in the destination state compared to the health status evaluation index value in the source state, A planning device equipped with the following features.

[0215] (Note 2) The path search means uses the health status evaluation index value in the destination state as an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state. The planned equipment as described in Appendix 1.

[0216] (Note 3) The system includes a target state setting means that determines the target state in the plan based on the health status evaluation index value for each state, The path search means searches for a path from the initial state in the plan to the target state determined by the target state setting means. The planned device as described in Appendix 1 or Appendix 2.

[0217] (Note 4) The path search means searches for paths from the initial state in the plan to individual states other than the initial state. Of the paths detected by the pathfinding means, the path from the initial state to the state set as the target state is used as the path from the initial state to the target state in the plan. The planned device as described in Appendix 1 or Appendix 2.

[0218] (Note 5) The means for acquiring the health status evaluation index value acquires the health status evaluation index value for each state that can be reached within a predetermined number of state transitions from the initial state in the plan. The planned device described in any one of the appendices 1 to 4.

[0219] (Note 6) The path search means searches for a path from the initial state to the target state in the plan using an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each state transition, which is obtained by weighting and summing an evaluation value for each state transition, which is an evaluation value for a single state transition from the source state to the destination state, where the evaluation value for the easier it is to realize the destination state is a good evaluation, and an evaluation value for the better the health evaluation index value in the destination state is compared to the health evaluation index value in the source state. A planned device as described in any one of the appendices 1 to 5.

[0220] (Note 7) The means for obtaining the health status evaluation index value calculates the health status evaluation index value using a model in which the parameters of the health status evaluation index, specifically the parameters that take values ​​according to a continuous distribution and the parameters that take values ​​according to a discrete distribution, have different hidden variables as parameters. A planned device as described in any one of the appendices 1 to 6.

[0221] (Note 8) The aforementioned condition is identified using the values ​​of one or more items that correlate with the health status assessment index value, and one of several time periods. The feasibility assessment index value acquisition means acquires the feasibility assessment index value for each value of one or more items that correlate with the health status assessment index value, and for each period. The means for acquiring the health status evaluation index value acquires the health status evaluation index value for each value of one or more items that correlate with the health status evaluation index value, and for each period. A planned device as described in any one of the appendices 1 through 7.

[0222] (Note 9) The health status evaluation index value acquisition means receives input values ​​for one or more items that identify the state and calculates the health status evaluation index value for each of the states using a trained model that outputs the health status evaluation index value for the identified state. A planned device as described in any one of the appendices 1 through 8.

[0223] (Note 10) From the measurement items for the subjects of the aforementioned plan, one or more items used to identify the aforementioned condition are selected based on the correlation between each measurement item and the health status evaluation index value. A planned device as described in any one of the appendices 1 through 9.

[0224] (Note 11) Computers For each state identified using the values ​​of one or more items that correlate with the health status evaluation index value, a feasibility evaluation index value is obtained that quantitatively indicates the ease with which each of the aforementioned states can be achieved. Obtain the health status evaluation index value for each of the aforementioned conditions, The system uses evaluation values ​​for each state transition, calculated using evaluation values ​​for one or more state transitions, to search for a path from the initial state to the target state in the plan. These evaluation values ​​indicate that the easier it is to achieve the target state, and that the better the health status index value in the target state is compared to the health status index value in the source state, the better the evaluation. A planning method that includes this.

[0225] (Note 12) In searching the aforementioned path, the computer uses the health status evaluation index value in the destination state as an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state. The planning method described in Appendix 11.

[0226] (Note 13) The computer includes determining the target state in the plan based on the health status evaluation index value for each state, In searching for the aforementioned path, the computer searches for a path from the initial state in the plan to the determined target state. The planning method described in Appendix 11 or Appendix 12.

[0227] (Note 14) In searching for the aforementioned path, the computer searches for a path from the initial state in the plan to each state other than the initial state. Of the paths detected during the path search, the path from the initial state to the state set as the target state is used as the path from the initial state to the target state in the plan. The planning method described in Appendix 11 or Appendix 12.

[0228] (Note 15) By obtaining the aforementioned health status evaluation index values, the computer obtains the aforementioned health status evaluation index values ​​for each state that can be reached within a predetermined number of state transitions from the initial state in the plan. The planning method described in any one of the appendices 11 to 14.

[0229] (Note 16) In exploring the aforementioned path, the computer uses an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each state transition, which is a weighted sum of an evaluation value for each state transition, where the evaluation value for a single state transition from the source state to the destination state is such that the easier it is to achieve the destination state, and an evaluation value for which the health status evaluation index value in the destination state is better than the health status evaluation index value in the source state. The computer then uses this evaluation value for a path consisting of one or more state transitions to explore the path from the initial state to the target state in the plan. The planning method described in any one of the appendices 11 to 15.

[0230] (Note 17) In obtaining the aforementioned health status evaluation index value, the computer calculates the health status evaluation index value using a model in which the parameters of the health status evaluation index, specifically the parameters that take values ​​according to a continuous distribution and the parameters that take values ​​according to a discrete distribution, have different hidden variables as parameters. The planning method described in any one of the appendices 11 to 16.

[0231] (Note 18) The aforementioned condition is identified using the values ​​of one or more items that correlate with the health status assessment index value, and one of several time periods. In obtaining the aforementioned feasibility assessment index values, the computer obtains the aforementioned feasibility assessment index values ​​for each value of one or more items that correlate with the health status assessment index value, and for each period. In obtaining the aforementioned health status evaluation index values, the computer obtains the health status evaluation index values ​​for each value of one or more items that correlate with the health status evaluation index values, and for each period. The planning method described in any one of the appendices 11 to 17.

[0232] (Note 19) In obtaining the aforementioned health status evaluation index value, the computer receives input values ​​for one or more items that identify the state and uses a trained model that outputs the health status evaluation index value for the identified state to calculate the health status evaluation index value for each of the aforementioned states. The planning method described in any one of the appendices 11 to 18.

[0233] (Note 20) The computer includes selecting one or more items from among the measurement items for the subject of the plan to be used to identify the state, based on the correlation between each measurement item and the health status evaluation index value. The planning method described in any one of the appendices 11 to 19.

[0234] (Note 21) On the computer, For each state identified using the values ​​of one or more items that correlate with the health status evaluation index value, a feasibility evaluation index value is obtained that quantitatively indicates the ease with which each of the aforementioned states can be achieved. Obtaining the health status evaluation index value for each of the aforementioned conditions, The process involves using evaluation values ​​for each individual state transition, calculated using evaluation values ​​for a path consisting of one or more state transitions, to explore a path from the initial state to the target state in a plan. These evaluation values ​​indicate that the easier it is to achieve the target state, the better the evaluation, and that the health status evaluation index value in the target state is better than the health status evaluation index value in the source state. A program that executes the command.

[0235] (Note 22) In searching for the aforementioned path, the computer is instructed to use the health status evaluation index value in the destination state as an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state. The program described in Appendix 21.

[0236] (Note 23) The computer is instructed to determine the target state in the plan based on the health status evaluation index value for each state. Searching for the aforementioned path involves causing the computer to search for a path from the initial state in the plan to the determined target state. The program described in Appendix 21 or Appendix 22.

[0237] (Note 24) In searching for the aforementioned path, the computer is instructed to search for paths from the initial state in the plan to each state other than the initial state. Of the paths detected during the path search, the path from the initial state to the state set as the target state is used as the path from the initial state to the target state in the plan. The program described in Appendix 21 or Appendix 22.

[0238] (Note 25) By obtaining the aforementioned health status evaluation index values, the computer is instructed to obtain the aforementioned health status evaluation index values ​​for each state that can be reached within a predetermined number of state transitions from the initial state in the plan. The program described in any one of the appendices 21 to 24.

[0239] (Note 26) In searching for the aforementioned path, the computer is instructed to search for a path from the initial state to the target state in the plan using an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each state transition, which is a weighted sum of an evaluation value for a single state transition from the source state to the destination state, where the evaluation value for the destination state is better the easier it is to achieve the destination state, and an evaluation value for the health evaluation index value in the destination state is better than the health evaluation index value in the source state. The program described in any one of the appendices 21 to 25.

[0240] (Note 27) In obtaining the aforementioned health status evaluation index value, the computer is instructed to calculate the health status evaluation index value using a model in which the parameters of the health status evaluation index, specifically the parameters that take values ​​following a continuous distribution and the parameters that take values ​​following a discrete distribution, have different hidden variables as parameters. The program described in any one of the appendices 21 to 26.

[0241] (Note 28) The aforementioned condition is identified using the values ​​of one or more items that correlate with the health status assessment index value, and one of several time periods. In obtaining the aforementioned feasibility assessment index values, the computer is instructed to obtain the aforementioned feasibility assessment index values ​​for each of the values ​​of one or more items that correlate with the health status assessment index values, and for each period. In obtaining the aforementioned health status evaluation index values, the computer is instructed to obtain the aforementioned health status evaluation index values ​​for each of the values ​​of one or more items that correlate with the health status evaluation index values, and for each period. The program described in any one of the appendices 21 to 27.

[0242] (Note 29) In obtaining the aforementioned health status evaluation index values, the computer is instructed to calculate the health status evaluation index values ​​for each of the aforementioned states using a trained model that receives input values ​​for one or more items that identify the state and outputs the health status evaluation index values ​​for the identified state. The program described in any one of the appendices 21 to 28.

[0243] (Note 30) The computer is instructed to select one or more items from the measurement items for the subject of the plan that are used to identify the state, based on the correlation between each measurement item and the health status evaluation index value. The program described in any one of the appendices 21 to 29. [Explanation of Symbols]

[0244] 100, 610 Planning device 110 Communications Department 120 Display section 130 Operation Input Section 180 Storage section 190 Processing Unit 191 Data Acquisition Unit 192, 612 Health Status Evaluation Index Value Acquisition Unit 193, 611 Feasibility Evaluation Index Value Acquisition Unit 194 Goal Setting Department 195, 613 Route search unit

Claims

1. A feasibility evaluation index value acquisition means for acquiring a feasibility evaluation index value that quantitatively indicates the ease of achieving each of the states identified using the values ​​of one or more items that correlate with the health status evaluation index value, A means for acquiring health status evaluation index values ​​for acquiring the health status evaluation index values ​​for each of the aforementioned states, A path search means for searching a path from an initial state to a target state in a plan, using an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each individual state transition, wherein the evaluation of a single state transition from the source state to the destination state is such that the easier it is to realize the destination state, the better the evaluation, and the better the health status evaluation index value in the destination state compared to the health status evaluation index value in the source state, A planning device equipped with the following features.

2. The path search means uses the health status evaluation index value in the destination state as an evaluation value that indicates a better evaluation when the health status evaluation index value in the destination state shows a better evaluation than the health status evaluation index value in the source state. The planning apparatus according to claim 1.

3. The system includes a target state setting means that determines the target state in the plan based on the health status evaluation index value for each state, The path search means searches for a path from the initial state in the plan to the target state determined by the target state setting means. The planning apparatus according to claim 1.

4. The path search means searches for paths from the initial state in the plan to individual states other than the initial state. Of the paths detected by the pathfinding means, the path from the initial state to the state set as the target state is used as the path from the initial state to the target state in the plan. The planning apparatus according to claim 1.

5. The means for acquiring the health status evaluation index value acquires the health status evaluation index value for each state that can be reached within a predetermined number of state transitions from the initial state in the plan. The planning apparatus according to claim 1.

6. The path search means searches for a path from the initial state to the target state in the plan using an evaluation value for a path consisting of one or more state transitions, calculated using an evaluation value for each state transition, which is obtained by weighting and summing an evaluation value for each state transition, where an evaluation value indicating a better evaluation is given when the target state is easier to achieve, and an evaluation value indicating a better evaluation when the health evaluation index value in the target state shows a better evaluation than the health evaluation index value in the source state. The planning apparatus according to claim 1.

7. The means for obtaining the health status evaluation index value calculates the health status evaluation index value using a model in which the parameters of the health status evaluation index, specifically the parameters that take values ​​according to a continuous distribution and the parameters that take values ​​according to a discrete distribution, have different hidden variables as parameters. The planning apparatus according to claim 1.

8. The aforementioned condition is identified using the values ​​of one or more items that correlate with the health status assessment index value, and one of several time periods. The feasibility assessment index value acquisition means acquires the feasibility assessment index value for each value of one or more items that correlate with the health status assessment index value, and for each period. The means for acquiring the health status evaluation index value acquires the health status evaluation index value for each value of one or more items that correlate with the health status evaluation index value, and for each period. The planning apparatus according to claim 1.

9. Computers For each state identified using the values ​​of one or more items that correlate with the health status evaluation index value, a feasibility evaluation index value is obtained that quantitatively indicates the ease with which each of the aforementioned states can be achieved. Obtain the health status evaluation index value for each of the aforementioned conditions, The system uses evaluation values ​​for each state transition, calculated using evaluation values ​​for one or more state transitions, to search for a path from the initial state to the target state in the plan. These evaluation values ​​indicate that the easier it is to achieve the target state, and that the better the health status index value in the target state is compared to the health status index value in the source state, the better the evaluation. A planning method that includes this.

10. On the computer, To obtain feasibility evaluation index values ​​that quantitatively show the ease of achieving each of the states identified using the values ​​of one or more items that correlate with the health status evaluation index values, Obtaining the health status evaluation index value for each of the aforementioned conditions, The process involves using evaluation values ​​for each individual state transition, calculated using evaluation values ​​for a path consisting of one or more state transitions, to explore a path from the initial state to the target state in a plan. These evaluation values ​​indicate that the easier it is to achieve the target state, the better the evaluation, and that the health status evaluation index value in the target state is better than the health status evaluation index value in the source state. A program that executes the command.

Images