Cognitive function state estimation model generation device, cognitive function state estimation model, and cognitive function state estimation device

Abstract

To provide a cognitive function state estimation model generation device, a cognitive function state estimation model and a cognitive function state estimation device for easily estimating a state of a human cognitive function with high accuracy.SOLUTION: A cognitive function state estimation model generation device 10 includes a feature amount generation part 13 for generating an input feature amount to be inputted to a cognitive function state estimation model, corresponding to cognitive function state information on the basis of oral function information of an object person in a time corresponding to an acquisition time of the cognitive function state information, a feature amount input part 14 for inputting the input feature amount to the cognitive function state estimation model, an update part 15 for updating a weighting coefficient constituting the cognitive function state estimation model on the basis of an error between output data outputted from the cognitive function state estimation model in accordance with the input of the input feature amount by the feature amount input part 14 and the cognitive function state information of learning data, and a model output part 16 for outputting the cognitive function state estimation model in which the weighting coefficient is updated by the update part 15.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a cognitive function state estimation model generation device, a cognitive function state estimation model, and a cognitive function state estimation device. [Background technology]

[0002] Using a model obtained by machine learning, diagnosis and evaluation of the mental and physical functions and conditions of a subject are performed based on various mental and physical data acquired from the subject. For example, Patent Document 1 discloses a system that outputs a diagnostic prediction by inputting data on nerve disorders collected by multiple sensors into a diagnostic model generated by machine learning. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2020-537579 Summary of the Invention [Problem to be solved by the invention]

[0004] It is known that there is a correlation between human oral functions such as the lips and tongue and cognitive functions such as memory, judgment, calculation, understanding, learning, and thinking. Based on this knowledge, training to improve oral functions has been recommended to prevent cognitive decline in the elderly and others, and there has been a need to estimate the state of cognitive function in order to understand and predict the effects of the training.

[0005] Therefore, the present invention has been made in consideration of the above problems, and has an object to easily and accurately estimate the state of a person's cognitive function. [Means for solving the problem]

[0006] In order to solve the above-described problems, a cognitive function state estimation model generation device according to one embodiment of the present invention is a cognitive function state estimation model generation device that generates, by machine learning, a cognitive function state estimation model that estimates a cognitive function state of a subject, wherein learning data used for the machine learning of the cognitive function state estimation model includes cognitive function state information of the subject and oral function information of the subject, the oral function information is information indicating a state of oral function related to the oral cavity of the subject, and the cognitive function state information is information indicating a cognitive function state of the subject that has been measured in advance, and the cognitive function state estimation model is constructed by predetermined machine learning, and is configured to estimate the cognitive function state in response to input of input feature amounts based on the oral function information. The cognitive function state estimation model generation device outputs state information and includes a feature generation unit that generates input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information, based on the oral function information of the subject at a time corresponding to the time when the cognitive function state information was acquired; a feature input unit that inputs the input features into the cognitive function state estimation model; an update unit that performs machine learning by updating the cognitive function state estimation model based on an error between output data output from the cognitive function state estimation model in response to input of the input features by the feature input unit and the cognitive function state information of the training data; and a model output unit that outputs the cognitive function state estimation model trained by the update unit.

[0007] According to the above embodiment, a cognitive function state estimation model is machine-learned using training data consisting of the subject's cognitive function state information and oral function information. The model is constructed by updating it based on the error between the cognitive function state information and output data output from the model in response to input of input features generated based on oral function information at a time corresponding to the time the cognitive function state information was acquired, thereby allowing the model to appropriately learn the correlation between oral function and the cognitive function state. Therefore, it is possible to obtain a cognitive function state estimation model that can accurately estimate the cognitive function state based on easily obtainable information such as oral function information. [Effects of the Invention]

[0008] It becomes possible to estimate the state of a person's cognitive function easily and with high accuracy. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing a functional configuration of a cognitive function state estimation model generation device according to an embodiment of the present invention. [Figure 2] 1 is a block diagram showing a functional configuration of a cognitive function state estimation device according to an embodiment of the present invention. [Figure 3] FIG. 2 is a hardware block diagram of a cognitive function state estimation model generation device and a cognitive function state estimation device. [Figure 4] FIG. 1 is a diagram illustrating a schematic configuration of a cognitive function state estimation model. [Figure 5] FIG. 2 is a diagram illustrating a schematic configuration of learning data. [Figure 6] FIG. 10 is a diagram illustrating an example of cognitive function state information. [Figure 7] FIG. 2 is a diagram showing an example of a schematic configuration of oral cavity function information. [Figure 8] FIG. 10 is a diagram showing an example of the structure of oral cavity function information and the data included therein. [Figure 9] 10 is a flowchart showing the processing steps of a cognitive function state estimation model generation method in the cognitive function state estimation model generation device. [Figure 10] 10 is a flowchart showing the processing steps of a cognitive function state estimation method in the cognitive function state estimation device. [Figure 11] FIG. 10 is a diagram illustrating a configuration of a cognitive function state estimation model generation program. [Figure 12] FIG. 10 is a diagram showing the configuration of a cognitive function state estimation program. DETAILED DESCRIPTION OF THE INVENTION

[0010] Embodiments of a cognitive function state estimation model generation device, a cognitive function state estimation device, and a cognitive function state estimation model according to the present invention will be described with reference to the drawings. Where possible, identical parts will be designated by the same reference numerals, and duplicated descriptions will be omitted.

[0011] The cognitive function state estimation model of this embodiment is a model constructed by machine learning to cause a computer to function and estimate the cognitive function state of a subject. The cognitive function state estimation model is capable of estimating the cognitive function state based on the subject's oral function information, and the type of model is not limited as long as it is a model generated by machine learning. The cognitive function state estimation model may, for example, include a neural network and output cognitive function state information by a predetermined calculation using a weighting coefficient for an input feature amount generated and input based on the oral function information. The oral function information is information indicating the state of oral function related to the subject's oral cavity. The cognitive function state information is information indicating the cognitive function state of the subject, and is expressed as scores of predetermined items indicating cognitive function.

[0012] The cognitive function state estimation model generation device of this embodiment is a device that generates a cognitive function state estimation model by machine learning. The cognitive function state estimation device is a device that outputs cognitive function state information in response to input oral function information, using the cognitive function state estimation model constructed by machine learning.

[0013] Fig. 1 is a diagram showing the functional configuration of a cognitive function state estimation model generation device according to this embodiment. As shown in Fig. 1, a cognitive function state estimation model generation device 10 functionally includes an oral function information acquisition unit 11, a cognitive function state information acquisition unit 12, a feature amount generation unit 13, a feature amount input unit 14, an update unit 15, and a model output unit 16. These functional units 11 to 16 may be configured in a single device or may be distributed across multiple devices.

[0014] Furthermore, the cognitive function state estimation model generation device 10 is configured to be able to access storage means such as a training data storage unit 30 and a model storage unit 40. The training data storage unit 30 and the model storage unit 40 may be configured within the cognitive function state estimation model generation device 10, or may be configured as separate devices external to the cognitive function state estimation model generation device 10 and accessible from the cognitive function state estimation model generation device 10, as shown in FIG.

[0015] The learning data storage unit 30 is a storage means that stores learning data used in machine learning of a cognitive function state estimation model, and can be configured with storage, memory, etc. The model storage unit 40 is a storage means that stores a cognitive function state estimation model that has been trained or is in the process of being trained, and can be configured with storage, memory, etc.

[0016] Fig. 2 is a diagram showing the functional configuration of a cognitive function state estimation device according to this embodiment. As shown in Fig. 2, the cognitive function state estimation device 20 functionally comprises an input data acquisition unit 21, an input feature generation unit 22, an input unit 23, and an output unit 24. These functional units 21 to 24 may be configured in a single device or may be distributed across multiple devices.

[0017] The cognitive function state estimation device 20 is also configured to be able to access a model storage unit 40 that stores a trained cognitive function state estimation model. The model storage unit 40 may be configured within the cognitive function state estimation device 20 or may be configured in a separate external device.

[0018] Furthermore, the cognitive function state estimation device 20 is configured to be able to access a storage means such as an oral function information storage unit 50. The oral function information storage unit 50 is a storage means that stores oral function information of a subject of cognitive function estimation. The oral function information storage unit 50 may be configured within the cognitive function state estimation device 20 or may be configured in a separate external device.

[0019] Furthermore, in this embodiment, an example is shown in which the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 are configured as separate devices (computers), but they may also be configured as an integrated device.

[0020] The block diagrams shown in FIGS. 1 and 2 show functional blocks. These functional blocks (components) are realized by any combination of at least one of hardware and software. Furthermore, there are no particular limitations on the method for realizing each functional block. That is, each functional block may be realized using a single device that is physically or logically coupled, or may be realized using two or more physically or logically separated devices that are connected directly or indirectly (for example, by wire, wirelessly, etc.) and these multiple devices. A functional block may also be realized by combining software with the single device or multiple devices.

[0021] Functions include, but are not limited to, judgment, determination, judgment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, election, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocation, mapping, and assignment. For example, a functional block (component) that performs transmission is called a transmitting unit or transmitter. As mentioned above, there are no particular limitations on how these functions are implemented.

[0022] For example, the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 according to an embodiment of the present invention may function as computers. Fig. 3 is a diagram showing an example of the hardware configuration of the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 according to this embodiment. The cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 may each be physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, etc.

[0023] In the following description, the term "apparatus" can be read as a circuit, a device, a unit, etc. The hardware configuration of the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 may be configured to include one or more of the devices shown in the drawings, or may be configured to exclude some of the devices.

[0024] Each function of the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 is realized by loading predetermined software (programs) onto hardware such as a processor 1001 and a memory 1002, causing the processor 1001 to perform calculations and control communication via a communication device 1004 and the reading and / or writing of data in the memory 1002 and storage 1003.

[0025] The processor 1001 controls the entire computer by running, for example, an operating system. The processor 1001 may be configured as a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. For example, the functional units 11 to 16, 21 to 24, etc. shown in FIGS. 1 and 2 may be realized by the processor 1001.

[0026] Furthermore, the processor 1001 reads programs (program codes), software modules, and data from the storage 1003 and / or the communication device 1004 into the memory 1002, and executes various processes in accordance with these. The programs used are those that cause a computer to execute at least some of the operations described in the above-described embodiments. For example, the functional units 11 to 16 and 21 to 24 of the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 may be implemented by a control program stored in the memory 1002 and running on the processor 1001. While the above-described various processes have been described as being executed by one processor 1001, they may also be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented on one or more chips. The programs may be transmitted from a network via a telecommunications line.

[0027] The memory 1002 is a computer-readable recording medium and may be composed of at least one of, for example, a read-only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a random access memory (RAM), etc. The memory 1002 may also be called a register, a cache, a main memory (primary storage device), etc. The memory 1002 can store executable programs (program codes), software modules, etc. for implementing the cognitive function state estimation model generation method and the cognitive function state estimation method according to one embodiment of the present invention.

[0028] Storage 1003 is a computer-readable recording medium, and may be, for example, at least one of an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray disc), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy disk, a magnetic strip, etc. Storage 1003 may also be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, a server, or other suitable medium including memory 1002 and / or storage 1003.

[0029] The communication device 1004 is hardware (transmission / reception device) for performing communication between computers via a wired and / or wireless network, and is also called, for example, a network device, a network controller, a network card, or a communication module.

[0030] The input device 1005 is an input device (for example, a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that receives input from the outside. The output device 1006 is an output device (for example, a display, a speaker, an LED lamp, etc.) that outputs to the outside. The input device 1005 and the output device 1006 may be integrated into one device (for example, a touch panel).

[0031] Furthermore, each device such as the processor 1001 and the memory 1002 is connected by a bus 1007 for communicating information. The bus 1007 may be configured as a single bus, or may be configured as different buses between the devices.

[0032] Furthermore, the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 may be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field programmable gate array (FPGA), and some or all of the functional blocks may be realized by the hardware. For example, the processor 1001 may be implemented by at least one of these pieces of hardware.

[0033] Fig. 4 is a diagram showing a schematic configuration of an example of a cognitive function state estimation model according to this embodiment. As shown in Fig. 4, the cognitive function state estimation model md includes a neural network, receives input feature values ​​fg generated based on oral function information of, performs calculations on the input feature values ​​fg input to the input layer of the neural network based on weighting coefficients and response functions corresponding to each layer, and outputs cognitive function state information cf from the output layer. The cognitive function state estimation model md may be a neural network composed of, for example, a multilayer perceptron, but the type of neural network constituting the cognitive function state estimation model md is not limited.

[0034] 1, the functional units of the cognitive function state estimation model generation device 10 will be described. The oral cavity function information acquisition unit 11 acquires oral cavity function information of as learning data. The cognitive function state information acquisition unit 12 acquires cognitive function state information cf as learning data.

[0035] Fig. 5 is a diagram showing a schematic configuration of the learning data. As shown in Fig. 5, the learning data ld used in the machine learning of the cognitive function state estimation model md includes oral function information of and cognitive function state information cf. The oral function information of and the cognitive function state information cf respectively constitute explanatory variables and target variables in the cognitive function state estimation model md.

[0036] In this embodiment, the learning data ld is pre-stored in the learning data storage unit 30, so the oral function information acquisition unit 11 acquires the oral function information of of the learning data ld by referring to the learning data storage unit 30. Furthermore, the cognitive function state information acquisition unit 12 acquires the cognitive function state information cf of the learning data ld by referring to the learning data storage unit 30.

[0037] 6 is a diagram showing an example of cognitive function state information. The cognitive function state information cf is information indicating the cognitive function state of a subject user that has been measured in advance, and includes a score related to at least one of memory, judgment, MCI (Mild Cognitive Impairment), dementia, and frailty, associated with identification information for identifying the user and date and time.

[0038] Specifically, the cognitive function state information cf may include scores indicating the degree of memory and judgment. The cognitive function state information cf may also include scores indicating the presence or absence of MCI, dementia, and frailty using binary values ​​of 1 or 0. Note that the date and time information included in the cognitive function state information cf means that the cognitive function state information is associated with a certain date and time, and the information for each included item does not necessarily have to be information measured at that date and time.

[0039] Fig. 7 is a diagram showing an example of the schematic configuration of oral function information. In the example shown in Fig. 7, the oral function information of includes state information is, state score ss, and body part state score sp. In the example shown in Fig. 7, the oral function information of includes three types of information: state information is, state score ss, and body part state score sp. However, in the cognitive function state estimation model generation device 10 and the cognitive function state estimation device 20 of this embodiment, the oral function information of only needs to include at least one of the state information is, state score ss, and body part state score sp.

[0040] 8 is a diagram showing an example of the structure of oral function information and the data included therein. As shown in Fig. 8, the oral function information of indicates the state of oral function related to the oral cavity of the subject user, and includes identification information for identifying the user, status information is associated with the date and time, a status score ss, and a site-specific status score sp.

[0041] The status information is is measurement data obtained by a predetermined measurement of oral cavity functions of the user (subject). The status information is is obtained by a predetermined analysis based on a facial image of the user and a voice uttered by the user. In an example of the status information is shown in FIG. 8, the status information is includes the number of pronunciations, clarity, and duration, but the content of the status information is is not limited to these examples.

[0042] The pronunciation count in the status information is is obtained, for example, by having the user pronounce predetermined sounds such as "pa," "ta," "ka," and "ra" and measuring the number of pronunciations within a predetermined time. The clarity in the status information is data indicating the degree of clarity of the predetermined sounds that the user has pronounced, and is obtained by a predetermined audio analysis of the vocalizations. The maintenance time in the status information is the time during which predetermined tongue and lip movements, etc., are continuously maintained by the user, and is obtained by image analysis of the user's facial image.

[0043] The condition score ss is a score indicating the degree of oral function, generated based on the condition information is. The condition score ss is obtained, for example, by referring to a table that previously associates various measurement data acquired as the condition information is with scores indicating the degree of oral function. The table referenced here may be a preset table, or may be one that is appropriately updated based on the distribution of previously acquired condition scores ss. In the example of the condition score ss shown in FIG. 8, the condition score ss includes scores for pronunciation, tongue movement, and facial expression, but the content of the condition score ss is not limited to these examples.

[0044] The site-specific condition score sp is a score obtained by aggregating the condition scores ss for each predetermined site related to the oral cavity. The site-specific condition score sp is calculated, for example, by pre-establishing an association between each condition score ss and a site, and aggregating the condition scores ss associated with each site using a predetermined aggregation method. The predetermined aggregation method may be, for example, calculating the sum or average of all the condition scores ss associated with one site, or may further include normalization of the calculated value. Furthermore, the average may be calculated using a preset weight.

[0045] The condition score sp for each body part includes at least one of a tongue score for the tongue, a lip score for the lips, a chin score for the chin, a cheek score for the cheeks, and an eye score for the eyes (including the eyebrows). In one example of the condition score sp for each body part shown in Fig. 8, the condition score sp for each body part includes the tongue score, the lip score, the chin score, the cheek score, and the eye score, and further includes an overall score obtained by aggregating (for example, averaging) the tongue score, the lip score, the chin score, the cheek score, and the eye score.

[0046] In this way, information that appropriately represents the state of oral function, such as the body part condition score SP, the condition score SS, and the condition information IS, is used to train the cognitive function state estimation model MD, making it possible to obtain a cognitive function state estimation model MD that can estimate the cognitive function state with high accuracy.

[0047] Furthermore, the body part condition score SP used as oral function information OF appropriately represents the state of oral function, which has a strong correlation with the subject's cognitive function, making it possible to obtain a cognitive function state estimation model MD that can estimate the cognitive function state with high accuracy.

[0048] Referring again to Figure 1, the feature generation unit 13 generates an input feature fg to be input into the cognitive function state estimation model md in association with the cognitive function state information cf, based on the oral function information of the subject at the time corresponding to the time when the cognitive function state information cf was acquired.

[0049] Specifically, as an example of generating the input feature fg, the feature generator 13 may use the oral function information of the subject at a time corresponding to the time when the cognitive function state information cf was acquired as input feature fg to be input to the cognitive function state estimation model md in association with the cognitive function state information cf. By configuring the input feature fg in this manner, it is possible to construct a cognitive function state estimation model md that can estimate the cognitive function state at the time when the oral function information of was acquired.

[0050] As an example of generating the input feature fg, the feature generator 13 may generate a lag feature using the time when the cognitive function state information cf was acquired as the reference time as the input feature fg. That is, the feature generator 13 may use the oral function information of the subject at a time corresponding to a predetermined period before the time when the cognitive function state information cf was acquired as the input feature fg to be input into the cognitive function state estimation model md in association with the cognitive function state information cf. By configuring the input feature fg in this manner, it is possible to construct a cognitive function state estimation model md that can estimate the cognitive function state a predetermined period after the time when the oral function information of was acquired.

[0051] As an example of generating the input feature fg, the feature generator 13 may generate the input feature fg using a window function with the acquisition time of the cognitive function state information cf as a reference time. That is, the feature generator 13 may obtain an aggregated value by a predetermined aggregation method for each item of the oral function information of during a given period corresponding to the acquisition time of the cognitive function state information cf, and may use the aggregated value as an input feature to be input to the cognitive function state estimation model md in association with the cognitive function state information cf.

[0052] Specifically, the feature generation unit 13 may use one or more of the minimum, maximum, median, mean, and standard deviation of all values ​​of each item of the oral function information of during a given period corresponding to the time when the cognitive function state information cf was acquired as an aggregate value to be used as the input feature fg.

[0053] By configuring the input features fg in this manner, a cognitive function state estimation model md can be constructed that can estimate the cognitive function state at a time corresponding to the correspondence between the period when the oral function information of used to calculate the aggregate value was acquired and the time when the cognitive function state information cf was acquired.

[0054] In this way, by appropriately selecting the oral function information of used to generate the input feature fg based on the time of acquisition, it is possible to obtain a cognitive function state estimation model md that can estimate the cognitive function state at any time.

[0055] The feature input unit 14 inputs the input feature fg generated by the feature generation unit 13 to the cognitive function state estimation model md. Specifically, the feature input unit 14 converts the input feature fg into a vector representation using a predetermined and well-known technique, and inputs the input feature fg converted into the vector representation to the cognitive function state estimation model md.

[0056] The update unit 15 updates the cognitive function state estimation model MD based on the error between the output data output from the cognitive function state estimation model MD in response to the input of the input feature amount fg by the feature amount input unit 14 and the cognitive function state information CF of the training data LD. When the cognitive function state estimation model MD is configured to include a neural network, the update unit 15 updates the weighting coefficients constituting the cognitive function state estimation model MD based on the error between the output data output from the cognitive function state estimation model MD in response to the input of the input feature amount fg by the feature amount input unit 14 and the cognitive function state information CF of the training data LD. Specifically, the update unit 15 updates and optimizes the weighting coefficients based on the error between the output data and the cognitive function state information CF using a well-known technique such as backpropagation.

[0057] The model output unit 16 outputs the cognitive function state estimation model md whose weighting coefficients have been updated and optimized by the update unit 15. Specifically, the model output unit 16 stores, for example, the cognitive function state estimation model md that has been optimized by the update unit 15 and has been trained in the model storage unit 40 so that it can be used for estimating a cognitive function state.

[0058] The trained cognitive function state estimation model md can be regarded as a program module that is read or referenced by a computer and causes the computer to execute predetermined processes and realize predetermined functions.

[0059] That is, the trained cognitive function state estimation model md of this embodiment is used in a computer equipped with a CPU and a memory. Specifically, the CPU of the computer operates in accordance with instructions from the trained cognitive function state estimation model md stored in the memory to perform calculations on the input feature value fg input to the input layer of the neural network based on the trained weight coefficients and response functions corresponding to each layer, and to output an output value from the output layer.

[0060] Next, the functional units of the cognitive function state estimation device 20 will be described with reference to Fig. 2. The input data acquisition unit 21 acquires oral function information of the subject whose cognitive function state is to be estimated. Specifically, the input data acquisition unit 21 acquires oral function information stored in, for example, the oral function information storage unit 50. The oral function information acquired by the input data acquisition unit 21 has a format similar to that of the oral function information of described with reference to Figs. 7 and 8.

[0061] The input feature generation unit 22 generates input features fg to be input to the cognitive function state estimation model md based on the oral cavity function information at a time period corresponding to the time period of the subject whose cognitive function state is to be estimated.

[0062] Specifically, the input feature generation unit 22 generates input features fg based on the oral function information, similar to the feature generation unit 13 of the cognitive function state estimation model generation device 10. As described above, the cognitive function state estimation model md is configured to estimate a cognitive function state at any relative time corresponding to the time when the input oral function information was acquired, so the input feature generation unit 22 generates input features based on the oral function information at a relative time corresponding to the time when the cognitive function state is to be estimated.

[0063] The input unit 23 inputs the input feature fg generated by the input feature generation unit 22 to the cognitive function state estimation model md. Specifically, the input unit 23 converts the input feature fg into a vector representation using a predetermined and well-known technique, and inputs the input feature fg converted into the vector representation to the trained cognitive function state estimation model md.

[0064] The output unit 24 outputs the output data output from the cognitive function state estimation model md as cognitive function state information of the subject of cognitive function state estimation. The manner of output is not limited, and the output unit 24 may store the cognitive function state information, which is the output data output from the cognitive function state estimation model md, in a predetermined storage means, display it on a predetermined display device, or transmit it to a predetermined computer.

[0065] FIG. 9 is a flowchart showing the processing steps of the cognitive function state estimation model generation method in the cognitive function state estimation model generation device 10.

[0066] In step S1, the cognitive function state estimation model generation device 10 acquires training data. Specifically, the oral cavity function information acquisition unit 11 acquires oral cavity function information of as training data. The cognitive function state information acquisition unit 12 acquires cognitive function state information cf as training data.

[0067] In step S2, the feature generation unit 13 generates an input feature fg to be input into the cognitive function state estimation model md in association with the cognitive function state information cf, based on the oral function information of of the subject at the time corresponding to the time when the cognitive function state information cf was acquired.

[0068] In step S3, the feature input unit 14 inputs the input feature fg generated by the feature generation unit 13 in step S2 to the cognitive function state estimation model md.

[0069] In step S4, the update unit 15 updates the weighting coefficients constituting the cognitive function state estimation model md based on the error between the output data output from the cognitive function state estimation model md in response to the input of the input feature fg by the feature input unit 14 and the cognitive function state information cf of the learning data ld.

[0070] In step S5, the cognitive function state estimation model generation device 10 determines whether to terminate learning of the cognitive function state estimation model md. The determination of termination of learning may be made based on the completion of machine learning using all learning data to be used for machine learning, the update of weighting coefficients and the like of the cognitive function state estimation model md, the convergence of optimization, etc.

[0071] If it is determined that the learning of the cognitive function state estimation model md should be terminated, the process proceeds to step S6. On the other hand, if it is not determined that the learning of the cognitive function state estimation model md should be terminated, the process returns to step S1. Alternatively, if the acquisition of all learning data and the generation of input features have been completed, the process may return to step S3.

[0072] In step S6, the model output unit 16 outputs the cognitive function state estimation model md in which the weighting coefficients have been updated and optimized.

[0073] FIG. 10 is a flowchart showing the processing steps of the cognitive function state estimation method using the trained cognitive function state estimation model md in the cognitive function state estimation device 20.

[0074] In step S11, the input data acquiring unit 21 acquires oral cavity function information of a subject whose cognitive function state is to be estimated.

[0075] In step S12, the input feature generation unit 22 generates input features fg to be input to the cognitive function state estimation model md based on oral function information at a period corresponding to the period of the subject whose cognitive function state is to be estimated.

[0076] In step S13, the input unit 23 inputs the input feature fg generated by the input feature generation unit 22 in step S12 to the cognitive function state estimation model md.

[0077] In step S14, the output unit 24 acquires the output data output from the cognitive function state estimation model md, and outputs the acquired output data as cognitive function state information of the subject whose cognitive function state is to be estimated.

[0078] Next, with reference to FIG. 11, a cognitive function state estimation model generation program for causing a computer to function as the cognitive function state estimation model generation device 10 of this embodiment will be described.

[0079] 11 is a diagram showing the configuration of a cognitive function state estimation model generation program. The cognitive function state estimation model generation program P1 includes a main module m10 that controls the cognitive function state estimation model generation process in the cognitive function state estimation model generation device 10, an oral cavity function information acquisition module m11, a cognitive function state information acquisition module m12, a feature generation module m13, a feature input module m14, an update module m15, and a model output module m16. The modules m11 to m16 implement the functions of the oral cavity function information acquisition unit 11, the cognitive function state information acquisition unit 12, the feature generation unit 13, the feature input unit 14, the update unit 15, and the model output unit 16.

[0080] The cognitive function state estimation model generation program P1 may be transmitted via a transmission medium such as a communication line, or may be stored in a recording medium M1 as shown in FIG. 11.

[0081] Next, with reference to FIG. 12, a cognitive function state estimating program for causing a computer to function as the cognitive function state estimating device 20 of this embodiment will be described.

[0082] 12 is a diagram showing the configuration of a cognitive function state estimation program. The cognitive function state estimation program P2 is configured to include a main module m20 that comprehensively controls the cognitive function state estimation process in the cognitive function state estimation device 20, an input data acquisition module m21, an input feature amount generation module m22, an input module m23, and an output module m24. The modules m21 to m24 realize the functions of the input data acquisition unit 21, the input feature amount generation unit 22, the input unit 23, and the output unit 24.

[0083] The cognitive function state estimation program P2 may be transmitted via a transmission medium such as a communication line, or may be stored in a recording medium M2 as shown in FIG.

[0084] According to the cognitive function state estimation model generation device 10, cognitive function state estimation model generation method, and cognitive function state estimation model generation program P1 of the present embodiment described above, a cognitive function state estimation model is machine-learned using training data consisting of the subject's cognitive function state information and oral function information. The cognitive function state estimation model is constructed by updating weighting coefficients based on the error between the cognitive function state information and output data output from the model in response to input feature values ​​generated based on oral function information at a time corresponding to the time the cognitive function state information was acquired, thereby allowing the model to appropriately learn the correlation between oral function and the cognitive function state. This makes it possible to obtain a cognitive function state estimation model that can accurately estimate the cognitive function state based on easily obtainable information such as oral function information.

[0085] In another embodiment of the cognitive function state estimation model generation device, the oral function information may include at least one of a body part-specific state score, a state score, and state information, the state information being measurement data obtained by a specified measurement of oral function related to the oral cavity of the subject, the state score being a score indicating the degree of oral function generated based on the state information, and the body part-specific state score being a score obtained by aggregating the state scores for each specified body part related to the oral cavity.

[0086] According to the above embodiment, information that appropriately represents the state of oral function, such as condition scores by region, condition scores, and condition information, is used to train the cognitive function state estimation model, making it possible to obtain a cognitive function state estimation model that can estimate the cognitive function state with high accuracy.

[0087] In addition, in another embodiment of the cognitive function state estimation model generation device, the oral function information may include at least a body part condition score, and the body part condition score may include at least one of a tongue score related to the tongue, a lip score related to the lips, a chin score related to the chin, a cheek score related to the cheeks, and an eye score related to the eyes.

[0088] According to the above embodiment, the body part condition scores used as oral function information appropriately represent the state of oral function that is highly correlated with the subject's cognitive function, making it possible to obtain a cognitive function state estimation model that can estimate the cognitive function state with high accuracy.

[0089] In addition, in another embodiment of the cognitive function state estimation model generation device, the feature generation unit may use the oral function information of the subject at a time corresponding to the time when the cognitive function state information was acquired as input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information, use the oral function information of the subject at a time corresponding to a predetermined period before the time when the cognitive function state information was acquired as input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information, or use the aggregated value obtained by a predetermined aggregation method of the oral function information for a given period corresponding to the time when the cognitive function state information was acquired as input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information.

[0090] According to the above aspect, by appropriately selecting the oral function information used to generate the input feature, it is possible to obtain a cognitive function state estimation model that can estimate the cognitive function state at any time.

[0091] In addition, in the cognitive function state estimation model generation device according to another aspect, the cognitive function state information may include a score related to at least one of memory, judgment, MCI, dementia, and frailty.

[0092] According to the above embodiment, the state of cognitive function of the subject can be estimated as the level of memory and judgment, and the presence or absence of MCI, dementia, and frailty.

[0093] In order to solve the above problem, a cognitive function state estimation model according to one embodiment of the present invention is a cognitive function state estimation model trained by machine learning for causing a computer to function and estimating the cognitive function state of a subject, wherein the learning data used for the machine learning of the cognitive function state estimation model comprises cognitive function state information of the subject and oral function information of the subject, the oral function information being information indicating the state of oral function related to the oral cavity of the subject, and the cognitive function state information being information indicating the cognitive function state of the subject that has been measured in advance, and the cognitive function state estimation model is constructed by machine learning that outputs cognitive function state information by performing a predetermined calculation on input features generated based on the oral function information, and updates the cognitive function state estimation model based on an error between the output data output in response to input of input features generated based on the oral function information of the subject at a time corresponding to the time the cognitive function state information was acquired and the cognitive function state information.

[0094] According to the above embodiment, a cognitive function state estimation model is constructed by machine learning using training data consisting of the subject's cognitive function state information and oral function information. The model is updated based on the error between the cognitive function state information and output data output from the model in response to input of input features generated based on oral function information at a time corresponding to the time the cognitive function state information was acquired, thereby constructing the cognitive function state estimation model. This allows the model to appropriately learn the correlation between oral function and the cognitive function state. Therefore, the cognitive function state can be estimated with high accuracy based on easily obtainable information such as oral function information.

[0095] In addition, in another form of cognitive function state estimation model, the oral function information includes at least one of a body part-specific state score, a state score, and state information, the state information is measurement data obtained by a specified measurement of oral function related to the oral cavity of the subject, the state score is a score indicating the degree of oral function generated based on the state information, and the body part-specific state score is a score obtained by aggregating the state scores for each specified body part related to the oral cavity.

[0096] According to the above embodiment, information that appropriately represents the state of oral function, such as condition scores by region, condition scores, and condition information, is used to train the cognitive function state estimation model, thereby realizing a cognitive function state estimation model that can estimate the cognitive function state with high accuracy.

[0097] In order to solve the above-described problems, a cognitive function state estimation device according to one embodiment of the present invention is a cognitive function state estimation device that estimates the cognitive function state of a subject using a cognitive function state estimation model constructed by machine learning, wherein learning data used for the machine learning of the cognitive function state estimation model includes cognitive function state information of the subject and oral function information of the subject, the oral function information is information indicating a state of oral function related to the oral cavity of the subject, and the cognitive function state information is information indicating the cognitive function state of the subject that has been measured in advance, and the cognitive function state estimation model is based on input features generated based on the oral function information of the subject at a time corresponding to a time when the cognitive function state information was acquired. The cognitive function state estimation device is constructed by machine learning that updates the cognitive function state estimation model based on an error between output data output as a result of a predetermined calculation on the cognitive function state and the cognitive function state information, and the cognitive function state estimation device includes an input feature generation unit that generates input features to be input into the cognitive function state estimation model based on oral function information of the subject whose cognitive function state is to be estimated, the oral function information being at a time corresponding to the time of the subject whose cognitive function state is to be estimated, an input unit that inputs the input features generated by the input feature generation unit into the cognitive function state estimation model, and an output unit that outputs the output data output from the cognitive function state estimation model as cognitive function state information of the subject whose cognitive function state is to be estimated.

[0098] According to the above embodiment, a cognitive function state estimation model constructed by machine learning using training data consisting of the subject's cognitive function state information and oral function information is used to estimate the cognitive function state. The model is constructed by updating the model based on the error between the cognitive function state information and output data output from the model in response to input of input feature values ​​generated based on oral function information at a time corresponding to the time the cognitive function state information was acquired. This allows the model to appropriately learn the correlation between oral function and the cognitive function state. Therefore, this model allows the cognitive function state to be estimated with high accuracy based on easily obtainable information such as oral function information.

[0099] In addition, in another form of cognitive function status estimation device, the oral function information may include at least one of a site-specific status score, a status score, and status information, where the status information is measurement data obtained by a specified measurement of oral function related to the oral cavity of the subject, the status score is a score indicating the degree of oral function generated based on the status information, and the site-specific status score is a score obtained by aggregating the status scores for each specified site related to the oral cavity.

[0100] According to the above aspect, the cognitive function state is estimated by a cognitive function state estimation model that is trained using information that appropriately expresses the state of oral function, such as the site-specific condition score, the condition score, and the condition information, as training data. Therefore, it is possible to estimate the cognitive function state with high accuracy.

[0101] In addition, in another embodiment of the cognitive function state estimation device, the input feature generation unit may use oral function information of the subject at a time corresponding to the target time of the subject whose cognitive function state is to be estimated as input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information, or may use oral function information of the subject at a time corresponding to a predetermined period before the target time of the subject whose cognitive function state is to be estimated as input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information, or may use an aggregated value obtained by a predetermined aggregation method of oral function information for a given period corresponding to the target time of the subject whose cognitive function state is to be estimated as input features to be input into the cognitive function state estimation model in correspondence with the cognitive function state information.

[0102] According to the above aspect, by appropriately selecting the oral function information used to generate the input feature amount, it becomes possible to estimate the cognitive function state at any time.

[0103] Although the present embodiment has been described in detail above, it is clear to those skilled in the art that the present embodiment is not limited to the embodiment described in this specification. The present embodiment can be implemented in modified and altered forms without departing from the spirit and scope of the present invention as defined by the claims. Therefore, the description in this specification is intended to be illustrative and does not have any limiting meaning on the present embodiment.

[0104] Each aspect / embodiment described herein may be applied to systems utilizing LTE (Long Term Evolution), LTE-Advanced (LTE-A), SUPER 3G, IMT-Advanced, 4G, 5G, FRA (Future Radio Access), W-CDMA (registered trademark), GSM (registered trademark), CDMA2000, UMB (Ultra Mobile Broadband), IEEE 802.11 (Wi-Fi), IEEE 802.16 (WiMAX), IEEE 802.20, UWB (Ultra-WideBand), Bluetooth (registered trademark), or other suitable systems and / or next generation systems enhanced thereon.

[0105] The order of the procedures, sequences, flowcharts, etc. of each aspect / embodiment described herein may be changed unless it is consistent. For example, the methods described herein present elements of various steps in an example order and are not limited to the particular order presented.

[0106] Input and output information may be stored in a specific location (for example, memory) or managed in a management table. Input and output information may be overwritten, updated, or added to. Output information may be deleted. Input information may be sent to another device.

[0107] The determination may be made based on a value represented by one bit (0 or 1), a Boolean value (true or false), or a numerical comparison (e.g., comparison with a predetermined value).

[0108] Each aspect / embodiment described in this specification may be used alone or in combination, or may be switched depending on the implementation. Furthermore, notification of predetermined information (e.g., notification that "X is true") is not limited to being done explicitly, but may be done implicitly (e.g., by not notifying the predetermined information).

[0109] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0110] Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

[0111] Software, instructions, etc. may also be transmitted or received over a transmission medium. For example, if software is transmitted from a website, server, or other remote source using wired technologies such as coaxial cable, fiber optic cable, twisted pair, and Digital Subscriber Line (DSL), and / or wireless technologies such as infrared, radio, and microwave, these wired and / or wireless technologies are included within the definition of transmission media.

[0112] The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0113] It should be noted that terms explained in this disclosure and / or terms necessary for understanding this specification may be replaced with terms having the same or similar meanings.

[0114] As used herein, the terms "system" and "network" are used interchangeably.

[0115] Furthermore, the information, parameters, etc. described in this specification may be expressed as absolute values, as relative values ​​from a predetermined value, or as corresponding other information.

[0116] As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Determining" and "determining" may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiring (e.g., searching in a table, database, or other data structure), ascertaining, and the like. "Determining" and "determining" may also include receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, accessing (e.g., accessing data in memory), and the like. Furthermore, "judgment" and "decision" can include regarding resolving, selecting, choosing, establishing, comparing, etc. as having been "judged" or "decided." In other words, "judgment" and "decision" can include regarding some action as having been "judged" or "decided." Furthermore, "judgment (decision)" can be interpreted as "assuming," "expecting," "considering," etc.

[0117] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly specified otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0118] When designations such as "first," "second," etc. are used herein, any reference to such elements does not generally limit the quantity or order of those elements. These designations may be used herein as a convenient method of distinguishing between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed therein or that the first element must precede the second element in some way.

[0119] To the extent that the terms "include," "including," and variations thereof are used herein or in the claims, these terms are intended to be inclusive, similar to the term "comprising." Furthermore, the term "or," as used herein or in the claims, is not intended to be an exclusive or.

[0120] In this specification, plural devices are also included unless the context or the technology clearly indicates that only one device is present.

[0121] Throughout this disclosure, the plural is intended to be included unless the singular is clearly indicated by the context. [Explanation of symbols]

[0122] 10...cognitive function state estimation model generation device, 11...oral function information acquisition unit, 12...cognitive function state information acquisition unit, 13...feature generation unit, 14...feature input unit, 15...update unit, 16...model output unit, 20...cognitive function state estimation device, 21...input data acquisition unit, 22...input feature generation unit, 23...input unit, 24...output unit, 30...learning data storage unit, 30...learning data storage unit, 40...model storage unit, 50...oral function information storage unit, M1...recording medium, m10...main module, m11...oral function information acquisition Module, m12...cognitive function state information acquisition module, m13...feature generation module, m14...feature input module, m15...update module, m16...model output module, M2...recording medium, m20...main module, m21...input data acquisition module, m22...input feature generation module, m23...input module, m24...output module, md...cognitive function state estimation model, P1...cognitive function state estimation model generation program, P2...cognitive function state estimation program.

Claims

1. A cognitive function state estimation model generation device that generates a cognitive function state estimation model that estimates a cognitive function state of a subject by machine learning, learning data used in machine learning of the cognitive function state estimation model includes cognitive function state information of the subject and oral function information of the subject; The oral function information is information indicating the state of oral function related to the oral cavity of the subject, the cognitive function state information is information indicating a cognitive function state of the subject that has been measured in advance, the cognitive function state estimation model is constructed by predetermined machine learning, and outputs cognitive function state information in response to input of input feature amounts based on the oral function information; The cognitive function state estimation model generation device includes: a feature generation unit that generates, based on the oral function information of the subject associated with a time period corresponding to an acquisition time period associated with the cognitive function state information, the input feature to be input into the cognitive function state estimation model in association with the cognitive function state information; a feature input unit that inputs the input feature into the cognitive function state estimation model; an update unit that performs machine learning by updating the cognitive function state estimation model based on an error between output data output from the cognitive function state estimation model in response to the input of the input feature amount by the feature amount input unit and the cognitive function state information of the training data; a model output unit that outputs the cognitive function state estimation model trained by the update unit; A cognitive function state estimation model generation device comprising:

2. The oral function information includes at least one of a site condition score, a condition score, and condition information, The condition information is measurement data obtained by a predetermined measurement of oral function related to the oral cavity of the subject, The condition score is a score indicating the degree of the oral function generated based on the condition information, The site-specific condition score is a score obtained by aggregating the condition scores for each predetermined site related to the oral cavity. The cognitive function state estimation model generating device according to claim 1 .

3. The oral function information includes at least the site condition score, The condition score by region includes at least one of a tongue score related to a tongue, a lip score related to a lip, a chin score related to a chin, and a cheek score related to a cheek, The cognitive function state estimation model generating device according to claim 2 .

4. The feature generation unit the oral function information of the subject at a time corresponding to the acquisition time of the cognitive function state information is used as an input feature to be input into the cognitive function state estimation model in association with the cognitive function state information; The oral function information of the subject at a time corresponding to a predetermined period before the time of acquisition of the cognitive function state information is used as an input feature to be input into the cognitive function state estimation model in association with the cognitive function state information; or a total value of the oral function information obtained by a predetermined aggregation method for a given period corresponding to the acquisition time of the cognitive function state information is used as an input feature to be input into the cognitive function state estimation model in association with the cognitive function state information; The cognitive function state estimation model generation device according to any one of claims 1 to 3.

5. The cognitive function status information includes a score on at least one of memory, judgment, MCI, dementia, and frailty. The cognitive function state estimation model generation device according to any one of claims 1 to 4.

6. A cognitive function state estimation model trained by machine learning for causing a computer to function and estimating a cognitive function state of a subject, learning data used in machine learning of the cognitive function state estimation model includes cognitive function state information of the subject and oral function information of the subject; The oral function information is information indicating the state of oral function related to the oral cavity of the subject, the cognitive function state information is information indicating a cognitive function state of the subject that has been measured in advance, The cognitive function state estimation model is outputting cognitive function state information by performing a predetermined calculation on the input feature amount generated based on the oral function information; The cognitive function state estimation model is constructed by machine learning, and the model is updated based on an error between output data output in response to input of the input feature amount generated based on the oral function information of the subject associated with a time period corresponding to an acquisition time period associated with the cognitive function state information and the cognitive function state information. Cognitive function status estimation model.

7. The oral function information includes at least one of a site condition score, a condition score, and condition information, The condition information is measurement data obtained by a predetermined measurement of oral function related to the oral cavity of the subject, The condition score is a score indicating the degree of the oral function generated based on the condition information, The site-specific condition score is a score obtained by aggregating the condition scores for each predetermined site related to the oral cavity. The cognitive function state estimation model according to claim 6 .

8. A cognitive function state estimation device that estimates a cognitive function state of a subject using a cognitive function state estimation model constructed by machine learning, learning data used in machine learning of the cognitive function state estimation model includes cognitive function state information of the subject and oral function information of the subject; The oral function information is information indicating the state of oral function related to the oral cavity of the subject, the cognitive function state information is information indicating a cognitive function state of the subject that has been measured in advance, the cognitive function state estimation model is constructed by machine learning, and the cognitive function state estimation model is updated based on an error between output data output as a result of a predetermined calculation on input features generated based on the oral function information of the subject associated with a time period corresponding to an acquisition time period associated with the cognitive function state information, and the cognitive function state information; The cognitive function state estimation device includes: an input feature generation unit that generates input features to be input to the cognitive function state estimation model based on oral function information of a subject whose cognitive function state is to be estimated, the oral function information being at a time corresponding to a time period of the subject whose cognitive function state is to be estimated; an input unit that inputs the input feature generated by the input feature generation unit into the cognitive function state estimation model; an output unit that outputs output data output from the cognitive function state estimation model as cognitive function state information of the subject of the cognitive function state estimation; A cognitive function state estimation device comprising:

9. The oral function information includes at least one of a site condition score, a condition score, and condition information, The condition information is measurement data obtained by a predetermined measurement of oral function related to the oral cavity of the subject, The condition score is a score indicating the degree of the oral function generated based on the condition information, The site-specific condition score is a score obtained by aggregating the condition scores for each predetermined site related to the oral cavity. The cognitive function state estimation device according to claim 8 .

10. The input feature generation unit the oral function information of the subject at a time corresponding to a target time of the subject whose cognitive function state is to be estimated is used as an input feature to be input into the cognitive function state estimation model in association with the cognitive function state information; The oral function information of the subject at a time corresponding to a predetermined period before the target time for estimating the cognitive function state is used as an input feature to be input into the cognitive function state estimation model in correspondence with the cognitive function state information; or a tabulation value of the oral function information for a given period corresponding to the target time period for estimating the cognitive function state, obtained by a predetermined tabulation method, is used as an input feature to be input into the cognitive function state estimation model in association with the cognitive function state information; The cognitive function state estimation device according to claim 8 or 9.

Images