Program, information processing device, and method

Abstract

Using data collected in an incomplete state, such as lifestyle data and clinical data of the subjects, an estimation model is used to predict the progression of a specified disease. [Solution] The server 20 of the disease progression prediction system 1 stores a database containing subject data about the subject (2021), and its functions include: acquiring subject data (S201); inputting subject data into an estimation model, which is a trained model that has learned past subject data and data indicating the state of a predetermined disease in the subject in the past, and causing it to output estimated data indicating the disease subtype and progression stage, which are the state of the subject for the predetermined disease (S202); acquiring the estimated data output by the estimation model (S203); and outputting the acquired estimated data (S204).

Description

Technical Field

[0001] The present disclosure relates to a program, an information processing apparatus, and a method.

Background Art

[0002] For the purpose of diagnosing a predetermined disease, various image data of a subject (patient) to be diagnosed, such as image data of CT examination, MRI examination, etc., are analyzed. Also, AI (artificial intelligence) technology is applied to the analysis of such image data.

[0003] Patent Document 1 discloses a technique for analyzing a brain morphological image and a brain functional image by an analysis model in a brain image analysis apparatus that outputs analysis data related to brain dysfunction, outputting analysis data related to brain dysfunction, and estimating cognitive impairment.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] By the way, a predetermined disease, such as dementia, is a disease that is pathologically and clinically diverse. In order to accurately estimate the state of the disease by a predetermined model (such as a learned model), it is necessary to individualize the model according to the attributes and situations of the subject. When trying to configure such an individualized model, multi-modal data such as life data and clinical data are required for each subject. However, since such data can only be obtained fragmentarily, it is difficult to collect longitudinal data for each subject, and it has been difficult to accurately estimate the state of the disease by a model.

[0006] Therefore, this disclosure describes a technology that enables the prediction of disease progression, such as cognitive function changes, using an estimation model, even when data is collected in an incomplete state, such as the lifestyle data and clinical data of the subject. [Means for solving the problem]

[0007] According to one embodiment of the present disclosure, a program is provided for a computer having a processor and memory to run on and estimate the state of a subject with respect to a predetermined disease. The memory stores a database containing subject data relating to the subject. The program causes the processor to perform the following steps: acquire subject data; input subject data into an estimation model, which is a trained model that has learned past subject data and data indicating the state of the subject with respect to a predetermined disease in the past, and output estimated data indicating the disease subtype and progression stage, which are the state of the subject with respect to the predetermined disease; acquire the estimated data output by the estimation model; and output the acquired estimated data. [Effects of the Invention]

[0008] According to this disclosure, subject data concerning a subject is input into an estimation model, and estimated data indicating the disease subtype and progression stage of a given disease for that subject is output. Therefore, it becomes possible to estimate the state of a given disease based on subject data concerning the subject, such as the subject's lifestyle data and clinical data. This makes it possible to predict the progression of a given disease using a model, even from data collected in an incomplete state, such as the subject's lifestyle data and clinical data. [Brief explanation of the drawing]

[0009] [Figure 1] This is a block diagram showing the overall configuration of the disease progression prediction system 1 according to an embodiment of the present disclosure. [Figure 2]Figure 1 is a block diagram showing the functional configuration of the terminal device 10. [Figure 3] This block diagram shows the functional configuration of server 20 in Figure 1. [Figure 4] Figure 3 shows an example of the data structure of the target database 2021. [Figure 5] Figure 3 shows an example of a graphical model of the estimated model 2022. [Figure 6] Figure 3 is a graph showing an example of the learning results using the estimated model 2022. [Figure 7] This flowchart shows an example of the learning process performed by the disease progression prediction system 1. [Figure 8] This flowchart shows an example of the flow of the estimation process performed by the disease progression prediction system 1. [Figure 9] This figure shows an example of the estimated result screen displayed on terminal device 10. [Figure 10] A block diagram showing the basic hardware configuration of Computer 90. [Modes for carrying out the invention]

[0010] The embodiments of this disclosure will be described below with reference to the drawings. In all the drawings illustrating the embodiments, common components are denoted by the same reference numerals, and repeated explanations are omitted. The following embodiments are not intended to unduly limit the content of this disclosure as described in the claims. Not all components shown in the embodiments are necessarily essential components of this disclosure. Also, each drawing is a schematic diagram and is not necessarily a strict illustration.

[0011] Furthermore, in the following description, "processor" refers to one or more processors. At least one processor is typically a microprocessor such as a CPU (Central Processing Unit), but may be another type of processor such as a GPU (Graphics Processing Unit). At least one processor may be single-core or multi-core.

[0012] Furthermore, at least one processor may be a broad-sense processor, such as a hardware circuit that performs some or all of the processing (e.g., an FPGA (Field-Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit)).

[0013] Furthermore, in the following explanation, we may use expressions such as "xxx database" to describe information that yields an output for a given input. This information can be data with any structure, or it can be a learning model such as a neural network that generates an output for a given input. Therefore, "xxx database" can be referred to as "xxx information."

[0014] Furthermore, in the following explanation, the table configurations for each database are examples only; one table may be divided into two or more tables, or all or part of two or more tables may constitute a single table.

[0015] Furthermore, in the following explanation, the subject of the process may sometimes be "program," but since a program is executed by a processor and performs defined processes using the memory and / or interface as appropriate, the subject of the process may also be the processor (or a device such as a controller that has that processor).

[0016] The program may be installed in a device such as a computer, or may be, for example, in a program distribution server or a computer-readable (e.g., non-transitory) recording medium. Also, in the following description, two or more programs may be realized as one program, or one program may be realized as two or more programs.

[0017] Also, in the following description, an identification number is used as identification information for various objects, but identification information of other types (e.g., an identifier including letters or symbols) other than the identification number may be adopted.

[0018] Also, in the following description, when describing without distinguishing between elements of the same type, reference signs (or common signs among the reference signs) are used, and when describing while distinguishing between elements of the same type, the identification numbers (or reference signs) of the elements may be used.

[0019] Also, in the following description, the control lines and information lines show those considered necessary for the description, and not necessarily all the control lines and information lines on the product are shown. All the components may be interconnected.

[0020] <Summary> Hereinafter, a disease progression prediction system according to the present disclosure will be described. The disease progression prediction system according to the present disclosure is a system used for estimating the state of a predetermined disease in a subject, for example, a disease such as dementia. This disease progression prediction system acquires multi-modal data such as life data and clinical data for each subject (patient) for whom the state is to be estimated, and based on that data, uses an estimation model to output estimation data indicating a disease subtype and a progression stage, which are the states of a predetermined disease of the subject. The disease progression prediction system according to the present disclosure is a system provided as a so-called SaaS (Software as a Service) as a web service by, for example, a cloud server or the like, and is configured to be accessible by a user through a predetermined authentication.

[0021] As mentioned above, certain diseases, such as dementia, are pathologically and clinically diverse in each individual. In order to accurately estimate the state of such diseases using a predetermined estimation model (such as a pre-trained model), it is necessary to individualize the estimation model according to the attributes and circumstances of the individual. To construct an estimation model for estimating the state of a disease in an individualized manner, multimodal data such as lifestyle data and clinical data are required for each individual. However, such data can only be obtained in fragments unless the individual is hospitalized or residing in various facilities, making it difficult to collect longitudinal data for each individual, and thus difficult to accurately estimate the state of a disease using an estimation model.

[0022] Therefore, the disease progression prediction system described in this disclosure is configured to acquire subject data such as lifestyle data and clinical data for each subject (patient) for whom the disease state is estimated, and to output estimated data showing the disease subtype and progression stage, which are the state of a given disease for that subject, using an estimation model based on that subject data.

[0023] Furthermore, the estimation model in the disease progression prediction system disclosed herein is composed of a probabilistic model (graphical model) that has been trained on past subject data and data indicating the state of a specified disease in past subjects. During the training of this estimation model, data indicating the interpretation of the past subject data and the data indicating the state of a specified disease in past subjects is output by a large-scale language model.

[0024] With this configuration, the disease progression prediction system described in this disclosure can be used to estimate the state of a given disease based on subject data, such as the subject's lifestyle data and clinical data. This makes it possible to predict the progression of a given disease using a model, even from data collected in an incomplete state, such as the subject's lifestyle data and clinical data. Furthermore, it makes it possible to personalize the estimation model for each subject.

[0025] <First Embodiment> The disease progression prediction system 1 according to an embodiment of this disclosure will be described below. In the following description, for example, when a terminal device 10 accesses a server 20, the server 20 responds with information for the terminal device 10 to generate a screen. The terminal device 10 generates and displays a screen based on the information received from the server 20.

[0026] <1 Overall configuration of disease progression prediction system 1> Figure 1 is a block diagram showing the overall configuration of the disease progression prediction system 1 according to Embodiment 1 of this disclosure. As shown in Figure 1, the disease progression prediction system 1 includes a plurality of terminal devices (in Figure 1, terminal devices 10A and 10B are shown; hereinafter, they may be collectively referred to as "terminal device 10"), a server 20, and an external server 30. The terminal devices 10, the server 20, and the external server 30 are connected to each other so as to be able to communicate with each other via a network 80. The network 80 is composed of a wired or wireless network. The network 80 includes, for example, 4G, 5G mobile communication systems, LTE (Long Term Evolution), and wireless networks that can connect to the internet via a predetermined access point (e.g., Wi-Fi®). When the network 80 is connected wirelessly, communication protocols include, for example, Z-Wave®, ZigBee®, and Bluetooth®. When the network is connected via a wired connection, the network also includes connections made directly by USB (Universal Serial Bus) cables, etc.

[0027] In this embodiment, server 20 is a web server (including a cloud server) and exchanges information with terminal device 10 via web pages. While terminal device 10 has a web browser installed for viewing web pages, a dedicated application for providing services from server 20 may also be installed and configured to allow viewing via this application. In this embodiment, the disease progression prediction system 1 is described as a configuration where terminal device 10 and server 20 are connected via a network 80, but it may also be configured on-premise using various standalone computer devices.

[0028] Terminal device 10 is a device operated by each user. Here, a user is someone who uses terminal device 10 to perform the function of the disease progression prediction system 1, which is to predict the progression of a disease such as dementia. This could be, for example, the person being treated, their family, or staff at a medical institution or nursing care facility. Terminal device 10 can be implemented as a stationary PC (Personal Computer), a laptop PC (Note PC), etc. Alternatively, terminal device 10 may be a mobile device such as a tablet compatible with a mobile communication system or a smartphone.

[0029] The terminal device 10 is connected to the server 20 via the network 80 in a communicative manner. The terminal device 10 is connected to the network 80 by communicating with communication equipment such as a wireless base station 81 that supports communication standards such as 4G, 5G, and LTE (Long Term Evolution), and a wireless LAN router 82 that supports wireless LAN (Local Area Network) standards such as IEEE (Institute of Electrical and Electronics Engineers) 802.11. As shown as terminal device 10B in Figure 1, the terminal device 10 includes a communication interface 12, an input device 13, an output device 14, a memory 15, a storage unit 16, and a processor 19.

[0030] The communication interface 12 is an interface for inputting and outputting signals so that the terminal device 10 can communicate with external devices. The input device 13 is an input device (for example, a keyboard, touch panel, touchpad, mouse, or other pointing device) for receiving input operations from the user. The output device 14 is an output device (display, speaker, etc.) for presenting information to the user. The memory 15 is for temporarily storing programs and data processed by programs, etc., and is a volatile memory such as DRAM (Dynamic Random Access Memory). The storage unit 16 is a storage device for saving data, such as flash memory or an HDD (Hard Disk Drive). The processor 19 is hardware for executing the instruction set written in the program, and is composed of an arithmetic unit, registers, peripheral circuits, etc.

[0031] Server 20 is a device that predicts the progression of diseases such as dementia and outputs the results. For each target person (patient) for disease progression prediction, Server 20 acquires target person data such as lifestyle data and clinical data. Based on the acquired target person data, Server 20 uses an estimation model to output estimated data indicating the disease subtype and progression stage, which are the state of the target person for a given disease. Furthermore, Server 20 trains the estimation model with past target person data and data indicating the state of the given disease in past target people.

[0032] Server 20 is a computer connected to network 80. Server 20 includes a communication interface 22, an input / output interface 23, memory 25, storage 26, and a processor 29.

[0033] Communication IF22 is an interface for inputting and outputting signals so that the server 20 can communicate with external devices. Input / Output IF23 functions as an interface to an input device for receiving input operations from the user and an output device for presenting information to the user. Memory 25 is for temporarily storing programs and data processed by programs, etc., and is a volatile memory such as DRAM (Dynamic Random Access Memory). Storage 26 is a storage device for saving data, such as flash memory or HDD (Hard Disk Drive). Processor 29 is hardware for executing the instruction set written in the program, and is composed of an arithmetic unit, registers, peripheral circuits, etc.

[0034] The external server 30 is, for example, a server device that provides services using a large-scale language model. The external server 30 outputs text data that is a structured document based on text data by providing the large-scale language model with a prompt containing text data. The external server 30 may be a large-scale language model trained on a large amount of text data, or a model obtained by transfer learning or fine-tuning the large-scale language model. Examples of large-scale language models include GPT-3 and GPT-4 developed by OpenAI, and Gemini developed by Google.

[0035] <1.1 Configuration of terminal device 10> Figure 2 is a block diagram showing the functional configuration of the terminal device 10 that constitutes the disease progression prediction system 1 of Embodiment 1. As shown in Figure 2, the terminal device 10 includes a plurality of antennas (antenna 111, antenna 112), wireless communication units corresponding to each antenna (first wireless communication unit 121, second wireless communication unit 122), an operation reception unit 130 (including a keyboard 131 and a mouse 132), a display 140, a storage unit 150, and a control unit 160. The terminal device 10 also has functions and configurations not specifically shown in Figure 2 (for example, a battery for maintaining power, a power supply circuit for controlling the supply of power from the battery to each circuit, etc.). As shown in Figure 2, each block included in the terminal device 10 is electrically connected by a bus or the like.

[0036] Antenna 111 radiates signals emitted by terminal device 10 as radio waves. Antenna 111 also receives radio waves from space and provides the received signals to first wireless communication unit 121.

[0037] Antenna 112 radiates signals emitted by terminal device 10 as radio waves. Antenna 112 also receives radio waves from space and provides the received signals to second wireless communication unit 122.

[0038] The first wireless communication unit 121 performs modulation and demodulation processing, etc., for the terminal device 10 to transmit and receive signals via the antenna 111 in order to communicate with other wireless devices. The second wireless communication unit 122 performs modulation and demodulation processing, etc., for the terminal device 10 to transmit and receive signals via the antenna 112 in order to communicate with other wireless devices. The first wireless communication unit 121 and the second wireless communication unit 122 are a communication module that includes a tuner, an RSSI (Received Signal Strength Indicator) calculation circuit, a CRC (Cyclic Redundancy Check) calculation circuit, a high-frequency circuit, etc. The first wireless communication unit 121 and the second wireless communication unit 122 perform modulation, demodulation, and frequency conversion of the wireless signals transmitted and received by the terminal device 10, and provide the received signal to the control unit 160.

[0039] The operation reception unit 130 has a mechanism for receiving user input operations. Specifically, the operation reception unit 130 includes a keyboard 131 and a mouse 132. The operation reception unit 130 may also be configured as a touchscreen that detects the user's contact position with the touch panel, for example, by using a capacitive touch panel.

[0040] The keyboard 131 accepts user input operations from the terminal device 10. The keyboard 131 is a device for character input and outputs the input character information as an input signal to the control unit 160.

[0041] The mouse 132 accepts user input operations from the terminal device 10. The mouse 132 is a pointing device for selecting objects displayed on the display 140, and outputs the selected position information on the screen and information indicating that a button is pressed as input signals to the control unit 160.

[0042] The display 140 displays data such as images, videos, and text in accordance with the control of the control unit 160. The display 140 is implemented, for example, by an LCD (Liquid Crystal Display) or an organic EL (Electro-Luminescence) display.

[0043] The storage unit 150 is composed of a memory 15, such as flash memory, and a storage unit 16, and stores data and programs used by the terminal device 10. In certain situations, the storage unit 150 stores user information 151.

[0044] User information 151 is information about a user who uses terminal device 10 to perform disease progression prediction, such as dementia, as a function of disease progression prediction system 1. User information may include information that identifies the user (user ID), the user's name, age, address, etc.

[0045] The control unit 160 is composed of, for example, a processor 19, and controls the operation of the terminal device 10 by reading a program stored in the memory unit 150 and executing the instructions contained in the program. The control unit 160 is, for example, an application that is pre-installed on the terminal device 10. By operating according to the program, the control unit 160 performs the functions of an input operation receiving unit 161, a transmitting / receiving unit 162, a notification control unit 163, and a data processing unit 164.

[0046] The input operation reception unit 161 processes input operations from the user to an input device such as a keyboard 131.

[0047] The transmitting / receiving unit 162 performs processing to enable the terminal device 10 to send and receive data with an external device such as a server 20 in accordance with a communication protocol.

[0048] The notification control unit 163 performs processing to present information to the user. The notification control unit 163 also performs processing such as displaying the display image on the display 140.

[0049] The data processing unit 164 performs calculations on the data received as input by the terminal device 10 according to the program and outputs the calculation results to memory or the like.

[0050] <1.2 Functional Configuration of Server 20> Figure 3 shows the functional configuration of the server 20 that constitutes the disease progression prediction system 1 of Embodiment 1. As shown in Figure 3, the server 20 functions as a communication unit 201, a storage unit 202, and a control unit 203.

[0051] The communications unit 201 performs processing to enable the server 20 to communicate with external devices.

[0052] The storage unit 202 stores data and programs used by the server 20. The storage unit 202 also stores the target database 2021 and the estimation model 2022, etc.

[0053] The Subject Database 2021 is a database for storing and maintaining various subject data related to subjects (patients) whose disease progression is predicted using the Disease Progression Prediction System 1. Subject Database 2021 includes, for example, subject data such as the subject's lifestyle data, clinical data from medical institutions, and data indicating the subject's disease subtype and progression stage. Further details will be described later.

[0054] Estimation Model 2022 is a pre-trained model used when predicting disease progression using the disease progression prediction system 1. Estimation Model 2022 consists of a probabilistic model (graphical model) that has been trained on past subject data and data indicating the state of a specified disease (dementia) in past subjects (disease subtype, progression stage). Here, the past subject data and data indicating the state of a specified disease in past subjects that are the subject of training for Estimation Model 2022 include data about the subject whose disease progression is being predicted and data about other people (patients). Further details will be described later.

[0055] The control unit 203 performs the functions shown in the reception control module 2031, transmission control module 2032, target data acquisition module 2033, target data input module 2034, estimation data acquisition module 2035, estimation data output module 2036, and estimation model learning module 2037 as various modules, by processing according to the program of the server 20's processor 29.

[0056] The receive control module 2031 controls the process by which the server 20 receives signals from external devices according to a communication protocol.

[0057] The transmission control module 2032 controls the process by which the server 20 transmits signals to external devices according to a communication protocol.

[0058] The subject data acquisition module 2033 controls the process of acquiring various subject data related to subjects (patients) for whom disease progression prediction is performed. For example, the user operates the terminal device 10 to input data identifying the subject (such as a user ID) and sends it to the server 20, so the subject data acquisition module 2033 receives the data sent from the terminal device 10 via the communication unit 201. At this time, the subject data acquisition module 2033 may acquire subject data such as lifestyle data from the user. Specifically, the subject data acquisition module 2033 may receive and acquire lifestyle data by answering questions in the form of a questionnaire, etc. Here, lifestyle data includes indicators related to lifestyle-related diseases (blood pressure, weight changes, eating habits, etc.), sleep patterns, physical activity levels, data on daily activities, and answers to questions in the form of a questionnaire, etc.

[0059] Furthermore, the subject data acquisition module 2033 may, for example, read the subject database 2021 for subject data transmitted when a user accesses it, and acquire stored subject lifestyle data, subject clinical data at medical institutions, data indicating the subject's disease subtype and progression stage, etc. Note that the disease subtype and progression stage do not need to be medically established classifications, and may include potential classifications, or may be classified according to the circumstances of the acquired data.

[0060] The target data acquisition module 2033 stores the target data acquired from user input in, for example, the target database 2021.

[0061] The subject data input module 2034 controls the process of inputting subject data acquired by the subject data acquisition module 2033 into the estimation model 2022 and outputting estimated data indicating the disease subtype and progression stage of a given disease (dementia) for that subject. The subject data input module 2034 is configured to use the estimation model 2022, which is a probabilistic model, to acquire estimated data indicating the disease subtype and progression stage for a given disease (dementia) for that subject as output data, based on the input subject data. Therefore, the subject data input module 2034 outputs estimated data indicating the disease subtype and progression stage by inputting subject data into the estimation model 2022.

[0062] Specifically, the subject data input module 2034 converts non-numerical data (e.g., text data) from the subject data acquired by the subject data acquisition module 2033 into numerical data based on predetermined criteria. This numerically converted subject data is then input into the estimation model 2022, which is a probability model. In addition, the subject data input module 2034 performs predetermined preprocessing on the subject data acquired by the subject data acquisition module 2033.

[0063] The estimation data acquisition module 2035 controls the process of acquiring the estimation data output by the estimation model 2022 in response to the input of subject data by the subject data input module 2034. The estimation data acquisition module 2035 acquires estimation data indicating the disease subtype and progression stage for a specified disease (dementia) corresponding to the subject data output by the estimation model 2022.

[0064] The estimated data acquired by the estimated data acquisition module 2035 includes, for example, numerical data on the state of a given disease (dementia), specifically the disease subtype and progression stage.

[0065] The estimated data acquisition module 2035 stores the acquired estimated data, for example, in the subject database 2021.

[0066] The estimated data output module 2036 controls the process of outputting the estimated data acquired by the estimated data acquisition module 2035. For example, the estimated data output module 2036 transmits estimated data indicating the disease subtype and progression stage for a predetermined disease (dementia) of a subject corresponding to the subject data to the terminal device 10 used by the user via the communication unit 201, and displays it on the display 140 of the terminal device 10. The estimated data output module 2036 may perform predetermined editing and processing on the estimated data indicating the disease subtype and progression stage before displaying it on the display 140 of the terminal device 10.

[0067] The estimation model learning module 2037 controls the process of training the estimation model 2022 with past subject data and data indicating the state of a given disease (dementia) in past subjects (disease subtype, progression stage). As described above, the estimation model 2022 is an estimation model that learns from past subject data and data indicating the state of dementia in past subjects using machine learning and performs estimation using probabilistic inference. Therefore, the estimation model learning module 2037 trains the estimation model with past subject data and data indicating the state of dementia in past subjects using machine learning.

[0068] The estimation model learning module 2037 may use, for example, human judgments or output by a large-scale language model for data showing the interpretation of past subject data and data showing the state of dementia in past subjects. The past subject data and data showing the state of a given disease in past subjects that the estimation model learning module 2037 learns include data about the subject themselves for whom disease progression prediction is performed, and data about other people (patients). Specifically, the estimation model learning module 2037 generates prompts for the large-scale language model to output past subject data and data showing the state of dementia in past subjects, and inputs them into the input field to access the external server 30 and provide them to the large-scale language model. The estimation model learning module 2037 then receives a response from the external server 30 and retrieves the output data. The data output from the large-scale language model may be text data (expressed in natural language) showing the interpretation of the state of dementia in past subjects, or it may be a functional type that shows changes in the data, specifically a sigmoid function, polynomial function, exponential function, segmented linear function, etc., as described later. In the case of a functional model that shows changes in data, users can more accurately understand the state of dementia by flexibly selecting and modifying that functional model. The estimation model learning module 2037 may also use text data output from the external server 30 to train the estimation model 2022.

[0069] <2 Data Structure> Figure 4 shows an example of the data structure of the target database 2021 shown in Figure 3.

[0070] As shown in Figure 4, each record in the target database 2021 includes the fields "Target ID", "Target", and "Target Data Details", etc.

[0071] The item "Subject ID" is information that identifies each subject (patient) who is the target of disease progression prediction system 1's disease status estimation.

[0072] The item "Subject" refers to the name, etc., of the subject (patient) whose disease status is estimated by the disease progression prediction system 1.

[0073] The item "Detailed Subject Data" refers to subject data relating to the subject (patient) who is the target of disease progression prediction system 1's disease status estimation, and specifically includes items such as "Date," "Vital Data," "Lifestyle Data," "Subtype," and "Stage."

[0074] The "Date" field represents the date on which the subject data was acquired. It indicates the date on which tests, etc., were conducted and the subject data shown in the "Vital Data" and "Lifestyle Data" fields was obtained.

[0075] The item "Vital Data" is data included in the subject data used for estimating the disease state by the disease progression prediction system 1, and is, for example, the subject's examination data and clinical data at a medical institution.

[0076] The item "Vital Data" may specifically contain the following data: ADAS-Cog: A scale used to assess cognitive function in Alzheimer's disease, which provides scores for word recall, oral language ability, auditory comprehension, etc. • PET Aβ: The result of a PET (Positron Emission Tomography) scan, which visualizes amyloid-beta and is used in the diagnosis of Alzheimer's disease, etc. • CSF p-tau: Measurement results of phosphorylated tau protein in cerebrospinal fluid.

[0077] The item "Lifestyle Data" is data included in the subject data used for estimating the disease state by the disease progression prediction system 1, and is data received, for example, through responses to questions in the form of a questionnaire.

[0078] The "Subtype" field is a disease subtype used by the Disease Progression Prediction System 1 to estimate the state of the disease, indicating a classification for different types of diseases. Examples of "Subtypes" in dementia include "Typical type (atrophy starting from the hippocampus and amygdala)," "Cortical type (atrophy starting from the anterior cingulate cortex and insula)," and "Subcortical type (atrophy starting from the globus pallidus and striatum)." The "Subtype" field stores numerical data (discrete values) assigned to each subtype.

[0079] The "Stage" field is a numerical value indicating the progression of the disease, used by the disease progression prediction system 1 to estimate the state of the disease. An example of the "Subtype" field is the storage of numerical data (continuous values) indicating the mild to severe stages of dementia.

[0080] The target data acquisition module 2033 on server 20 adds a record to the "Target Data Details" field in the target database 2021 when it acquires target data.

[0081] Figure 5 shows an example of a graphical model of the estimation model 2022 shown in Figure 3. The example shown in Figure 5 is a type of graphical model that visually describes the structure of estimation model 2022 in an easy-to-understand way, and is a Bayesian network that visualizes the dependencies between multiple variables using probabilistic inference. The subject data input module 2034 uses the estimation model 2022 shown in Figure 5 to output estimated data indicating the disease subtype and progression stage for a given disease (dementia) in the subject, using Bayesian estimation.

[0082] A Bayesian network is represented as a directed acyclic graph where variables are connected to other variables, with each variable being a node and the connections between nodes being links. In Figure 5, the elliptical nodes N1 to N13 of the Bayesian network correspond to the subject data, disease subtypes, progression stages, and estimated data that constitute the estimation model 2022, and hold random variables calculated by Bayesian estimation. In addition, the rectangular plate frames S1 to S4 in Figure 5, consisting of multiple nodes, indicate the number of times the value shown in the lower right is repeated. The example shown in Figure 5 is an example of training an estimation model defined with three disease subtypes and four progression stages using 665 data points.

[0083] The plate frame S1 shown in Figure 5 represents the model coefficients of the estimated model 2022. Nodes N1, N2, and N5 represent semi-normal distributions (distributions obtained by folding a normal distribution with mean 0 on the negative side), and node N4 represents a normal distribution. The plate frame S2 represents the disease subtype and progression stage of the estimated model 2022. Node N3, representing the disease subtype, represents a Dirichlet distribution (prior distribution to the probability vector), node N6, representing the progression stage, represents a uniform distribution, and node N7 represents a categorical distribution (discrete distribution). The plate frame S3 is a function representing the operations of nodes N4 to N6, and node N8 represents a deterministic function. The plate frame S4 ​​represents the estimates from the estimated model 2022. Node N9, representing the estimates, represents a deterministic function representing the operations of nodes N7 and N8, node N11 represents a normal distribution, node N12 represents actual data, and node N13 represents missing data.

[0084] The estimated model 2022 can be represented, for example, by the following sigmoid function formula. The estimated model 2022 may also be represented by a polynomial function, an exponential function, or a segmented linear function.

number

[0085] The subject data input module 2034 outputs estimated data using the estimation model 2022, which is a probabilistic model. The algorithm used at this time is designed to simultaneously estimate disease subtype and progression stage from subject data, which are multiple markers. The purpose of this algorithm is to estimate pseudo-longitudinal time-series data from cross-sectional subject data by introducing a latent time axis using estimation model 2022 and reproducing the time-series structure.

[0086] Figure 6 is a graph showing an example of the learning results using the estimated model 2022 from Figure 3. The graphs in Figure 6 show 3D normalized data when ADAS-Cog, PET Aβ, and CSF p-tau are used as markers, and are shown as graphs G1, G2, and G3 for each of the three subtypes.

[0087] The estimation model 2022 is assigned multiple disease subtypes and markers. Furthermore, for each point in graphs G1, G2, and G3, which graph the subject data, a stage latent variable X and a subtype latent variable z are assigned, and the observed values ​​are considered to be the values ​​at time X corresponding to z in estimation model 2022. By performing a batch estimation using MCMC (Markov Chain Monte Carlo method: an algorithm that samples probability distributions by creating a Markov chain with the desired probability distribution as the equilibrium distribution), missing values ​​are imputed during inference, making it possible to learn even from data containing noise or partial missing values, and to obtain the estimated stage and subtype probabilities for individual data points.

[0088] <3 operations> The learning process and estimation process (method) of the disease progression prediction system 1 according to the embodiment of this disclosure will be described below with reference to Figures 7 and 8.

[0089] Figure 7 is a flowchart showing an example of the learning process performed by the disease progression prediction system 1.

[0090] In step S101, the subject data acquisition module 2033 of the server 20 acquires subject data such as lifestyle data and clinical data for each subject (patient) whose disease state is estimated for disease progression prediction. In step S101, subject data may be acquired by user operation and transmitted from the terminal device 10, or by reading the subject database 2021.

[0091] In step S102, the subject data input module 2034 of the server 20 performs preprocessing on the subject data acquired in step S101. In step S102, for example, preprocessing is performed to quantify the subject data.

[0092] In step S103, the estimation model learning module 2037 of the server 20 defines the estimation model 2022 in order to learn from the subject data acquired in step S101. In step S103, the estimation model 2022 is defined, for example, as shown in Figure 5.

[0093] In step S104, the estimation model learning module 2037 of the server 20 trains the estimation model 2022, which was defined in step S103, with past subject data and data indicating the state of a given disease (dementia) in past subjects (disease subtype, progression stage). In step S104, for example, training is performed on the estimation model 2022 in order to perform probabilistic inference using Bayesian estimation.

[0094] As described above, the disease progression prediction system 1 acquires and preprocesses subject data such as lifestyle data and clinical data for each subject (patient) for whom disease status estimation is performed. It then learns from past subject data and data indicating the state of a given disease (dementia) in past subjects (disease subtype, progression stage), and performs estimation using probabilistic inference. As a result, by using the disease progression prediction system disclosed herein, even if there is missing data, it is imputed during the inference process, making it possible to learn even from data containing noise or partial missing data.

[0095] Figure 8 is a flowchart showing an example of the flow of estimation processing performed by the disease progression prediction system 1.

[0096] In step S201, the subject data acquisition module 2033 of the server 20 acquires subject data such as lifestyle data and clinical data for each subject (patient) whose disease state is estimated for disease progression prediction. In step S201, subject data may be acquired by user operation from the terminal device 10, or by reading the subject database 2021.

[0097] In step S202, the subject data input module 2034 of the server 20 inputs the subject data acquired in step S201 into the estimation model 2022 and outputs estimated data indicating the disease subtype and progression stage, which are the conditions of the subject with a predetermined disease (dementia).

[0098] In step S203, the estimated data acquisition module 2035 of the server 20 acquires the estimated data output by the estimation model 2022 in response to the input of subject data in step S202. In step S203, for example, estimated data indicating the disease subtype and progression stage is acquired for a predetermined disease (dementia) of a subject corresponding to the subject data.

[0099] In step S204, the estimated data output module 2036 of the server 20 outputs the estimated data acquired in step S203. In step S204, for example, estimated data indicating the disease subtype and progression stage for a predetermined disease (dementia) of a subject corresponding to the subject data is transmitted via the communication unit 201 to the terminal device 10 used by the user and displayed on the display 140 of the terminal device 10.

[0100] As described above, the disease progression prediction system 1 acquires subject data such as lifestyle data and clinical data for each subject (patient) for whom the disease state is estimated, and outputs estimated data (disease subtype, progression stage) indicating the state of a specified disease (dementia) for that subject using an estimation model. As a result, it is possible to estimate the state of a specified disease by using the disease progression prediction system disclosed herein.

[0101] <4. Screen Example> The following describes an example of the estimated results screen displayed on the terminal device 10 by the disease progression prediction system 1, with reference to Figure 9.

[0102] Figure 9 shows an example of the estimated result screen displayed on the terminal device 10. The example screen in Figure 9 shows a screen displaying estimated data indicating the disease subtype and progression stage, which are the estimated results for the subject (patient) for whom disease progression prediction is performed. This corresponds to step S204 in Figure 8.

[0103] As shown in Figure 9, the display data screen 1411 is displayed on the display 140 of the terminal device 10. This display data screen 1411 displays, for example, a subject display field 1412 that shows the subject, a subject data display field 1413 that shows subject data (examination data at a medical institution, lifestyle data) about the subject, and an estimated data display field 1414 that shows estimated data (disease subtype, progression stage) about the subject. By referring to this display data screen 1411, the user can understand the estimated data about the subject.

[0104] <Summary> As described above, according to this embodiment, subject data such as lifestyle data and clinical data are acquired for each subject (patient) for disease state estimation, and an estimation model is used based on that subject data to output estimated data (disease subtype, progression stage) indicating the state of a predetermined disease (dementia) for that subject. Therefore, by using the disease progression prediction system according to this disclosure, it becomes possible to predict changes in disease progression such as cognitive function and estimate the state of a predetermined disease, such as dementia, based on subject data about the subject, such as the subject's lifestyle data and clinical data. This makes it possible to perform disease progression prediction for a predetermined disease using a model, even from data collected in an incomplete state, such as the subject's lifestyle data and clinical data.

[0105] Furthermore, according to this embodiment, subject data such as lifestyle data and clinical data are acquired for each subject (patient) for disease state estimation. An estimation model is constructed using a probabilistic model (graphical model) that has learned from past subject data and data indicating the state of a predetermined disease in past subjects, and estimation is performed by probabilistic inference. As a result, by using the disease progression prediction system according to this disclosure, even if there is missing data, it will be imputed during the inference, making it possible to learn even from data containing noise or partial missing data. In addition, it becomes possible to individualize the estimation model for each subject.

[0106] <Basic Computer Hardware Configuration> Figure 10 is a block diagram showing the basic hardware configuration of computer 90. Computer 90 comprises at least a processor 901, main memory 902, auxiliary storage 903, and a communication interface IF991. These are electrically connected to each other by a communication bus 921.

[0107] The processor 901 is hardware for executing the instruction set written in a program. The processor 901 consists of an arithmetic unit, registers, peripheral circuits, etc.

[0108] Main memory 902 is used to temporarily store programs and data processed by programs, etc. For example, it is a volatile memory such as DRAM (Dynamic Random Access Memory).

[0109] Auxiliary storage device 903 refers to a storage device for saving data and programs. Examples include flash memory, HDD (Hard Disc Drive), magneto-optical disk, CD-ROM, DVD-ROM, and semiconductor memory.

[0110] The IF991 communication interface is an interface for inputting and outputting signals for communication with other computers via a network using wired or wireless communication standards. A network consists of various mobile communication systems, such as the internet, LANs, and wireless base stations. For example, a network includes 3G, 4G, and 5G mobile communication systems, LTE (Long Term Evolution), and wireless networks that can connect to the internet via designated access points (e.g., Wi-Fi®). When connecting wirelessly, communication protocols include, for example, Z-Wave®, ZigBee®, and Bluetooth®. When connecting via a wired connection, the network also includes connections made directly via USB (Universal Serial Bus) cables, etc.

[0111] Furthermore, by distributing all or part of each hardware configuration across multiple computers 90 and connecting them to each other via a network, a computer 90 can be virtually realized. Thus, the concept of computer 90 includes not only a computer 90 housed in a single enclosure or case, but also a virtualized computer system.

[0112] <Basic Functional Configuration of Computer 90> The functional configuration of the computer realized by the basic hardware configuration of computer 90 (Figure 10) is described below. The computer comprises at least one functional unit: a control unit, a memory unit, and a communication unit.

[0113] Furthermore, the functional units of computer 90 can also be realized by distributing all or part of each functional unit across multiple computers 90 interconnected via a network. The concept of computer 90 includes not only a single computer 90 but also a virtualized computer system.

[0114] The control unit is realized when the processor 901 reads various programs stored in the auxiliary storage device 903, loads them into the main memory device 902, and executes processing according to those programs. The control unit can realize various functional units that perform information processing depending on the type of program. In this way, the computer is realized as an information processing device that performs information processing.

[0115] The memory unit is implemented by the main memory 902 and the auxiliary memory 903. The memory unit stores data, various programs, and various databases. The processor 901 can also reserve memory areas corresponding to the memory unit in the main memory 902 or the auxiliary memory 903 according to the program. The control unit can also cause the processor 901 to perform operations such as adding, updating, and deleting data stored in the memory unit according to the various programs.

[0116] A database, specifically a relational database, is used to manage and link together tabular data sets called masters, which are structurally defined by rows and columns. In a database, tables are called tables, masters are called masters, the columns of tables are called columns, and the rows of tables are called records. In a relational database, relationships can be established and linked between tables and masters. Typically, each table and master has a primary key column to uniquely identify records, but setting a primary key column is not mandatory. The control unit can instruct the processor 901 to add, delete, or update records in specific tables and masters stored in the memory unit, according to various programs. Furthermore, by storing data, various programs, and various databases in the memory unit, the information processing device and information processing system related to this disclosure can be considered to have been manufactured.

[0117] Furthermore, the databases and masters in this disclosure may include any data structures (lists, dictionaries, associative arrays, objects, etc.) in which information is structurally defined. Data structures also include data that can be considered as data structures by combining data with functions, classes, methods, etc., written in any programming language.

[0118] The communication unit is implemented by the communication IF991. The communication unit provides the functionality to communicate with other computers 90 via the network. The communication unit can receive information transmitted from other computers 90 and input it to the control unit. The control unit can cause the processor 901 to perform information processing on the received information according to various programs. The communication unit can also transmit information output from the control unit to other computers 90.

[0119] Furthermore, each of the above-mentioned configurations, functions, processing units, processing means, etc., may be implemented in hardware, either partially or entirely, by designing them as integrated circuits, for example. The present invention can also be implemented by software program code that realizes the functions of the embodiment. In this case, a storage medium on which the program code is recorded is provided to a computer, and the processor of that computer reads the program code stored in the storage medium. In this case, the program code read from the storage medium itself realizes the functions of the embodiment described above, and the program code itself and the storage medium on which it is stored constitute the present invention. Examples of storage media used to supply such program code include flexible disks, CD-ROMs, DVD-ROMs, hard disks, SSDs, optical disks, magneto-optical disks, CD-Rs, magnetic tapes, non-volatile memory cards, ROMs, and the like.

[0120] Furthermore, the program code that implements the functions described in this embodiment can be implemented in a wide range of programming or scripting languages, such as assembler, C / C++, Perl, Shell, PHP, and Java (registered trademark).

[0121] Furthermore, the program code for the software that implements the functions of the embodiment may be distributed via a network and stored in a storage means such as a computer's hard disk or memory, or in a storage medium such as a CD-RW or CD-R, and the computer's processor may read and execute the program code stored in the storage means or storage medium.

[0122] The functions realized by the components described herein may be implemented in a circuit or processing circuitry, including general-purpose processors, application-specific processors, integrated circuits, ASICs (Application Specific Integrated Circuits), CPUs (A Central Processing Units), conventional circuits, and / or combinations thereof, programmed to realize the functions described herein. A processor is considered to be a circuit or processing circuitry, including transistors and other circuits. A processor may be a programmed processor that executes a program stored in memory. In this specification, circuitry, unit, and means are hardware programmed to perform or execute the functions described herein. Such hardware may be any hardware disclosed herein, or any hardware known to be programmed to perform or execute the functions described herein. If the hardware is a processor that is considered to be a type of circuitry, then the circuitry, means, or unit is a combination of hardware and software used to constitute the hardware and / or processor.

[0123] While embodiments relating to this disclosure have been described above, these can be implemented in various other forms and can be carried out with various omissions, substitutions, and modifications. These embodiments and variations, as well as those with omissions, substitutions, and modifications, are included within the technical scope of the claims and their equivalents.

[0124] <Note> The details described in each of the above embodiments are noted below.

[0125] (Note 1) A program to be executed on a computer equipped with a processor 29 and memory 25 for estimating the state of a subject with respect to a predetermined disease, wherein the memory 25 stores a database containing subject data relating to the subject (2021), and the program causes the processor 29 to perform the steps of: acquiring subject data (S201); inputting subject data into an estimation model, which is a trained model that has learned past subject data and data indicating the state of the subject with respect to a predetermined disease in the past, and causing it to output estimated data indicating the disease subtype and progression stage, which are the state of the subject with respect to the predetermined disease (S202); acquiring the estimated data output by the estimation model (S203); and outputting the acquired estimated data (S204).

[0126] (Note 2) The program further includes the step (S104) of causing the estimation model to learn past subject data and data indicating the state of a given disease in past subjects, as described in (Note 1).

[0127] (Note 3) The program further includes the steps of generating a prompt that causes a large language model to output data indicating an interpretation of either or both of the past subject data or data indicating the state of a given disease in past subjects, and providing the generated prompt to the large language model and receiving a response from the large language model, and training the estimation model, thereby training the output received from the large language model, as described in (Note 2).

[0128] (Note 4) The estimation model is a program described in any of (Note 1) to (Note 3), consisting of a probabilistic model that has learned from past subject data and data indicating the state of a given disease in past subjects.

[0129] (Note 5) The estimation model is the program described in (Note 4), which consists of a graphical model in which disease subtypes and progression stages related to a given disease are defined in advance.

[0130] (Note 6) The graphical model is the program described in (Note 5), in which model coefficients are determined based on learned past subject data and data showing the state of a given disease in past subjects.

[0131] (Note 7) The estimation model is the program described in (Note 4), which outputs estimated data indicating the disease subtype and progression stage of the dementia condition of the subject.

[0132] (Note 8) The participant data includes either or both of the participant's lifestyle data and / or clinical data of the participant at a medical institution, as described in (Note 1).

[0133] (Note 9) An information processing device comprising a control unit 203 and a memory 25 (storage unit 202) for estimating the state of a subject with respect to a predetermined disease, wherein the memory 25 stores a database storing subject data relating to the subject (2021), and the control unit 203 performs the steps of acquiring subject data (S201), inputting subject data to an estimation model which is a trained model that has learned past subject data and data indicating the state of the subject with respect to a predetermined disease in the past, and causing it to output estimated data indicating the disease subtype and progression stage which are the state of the subject with respect to a predetermined disease (S202), acquiring the estimated data output by the estimation model (S203), and outputting the acquired estimated data (S204).

[0134] (Note 10) A method for estimating the state of a subject with respect to a predetermined disease, which is executed by a computer comprising a processor 29 and a memory 25, wherein the memory 25 stores a database containing subject data relating to the subject (2021), and the method is a method in which the processor 29 performs the steps of: acquiring subject data (S201); inputting the subject data into an estimation model, which is a trained model that has learned past subject data and data indicating the state of a subject with respect to a predetermined disease in the past, and causing it to output estimated data indicating the disease subtype and progression stage, which are the state of the subject with respect to a predetermined disease (S202); acquiring the estimated data output by the estimation model (S203); and outputting the acquired estimated data (S204). [Explanation of Symbols]

[0135] 1: Disease progression prediction system 10: Terminal device 10A: Terminal device 10B: Terminal device 13: Input device 14: Output device 15: Memory 16: Storage section 19: Processor 20: Server 25: Memory 26: Storage 29: Processor 30: External Server 80: Network 81: Wireless base station 82: Wireless LAN router 90: Computer 111: Antenna 112: Antenna 121: First Radio Communication Unit 122: Second Wireless Communication Section 130: Operation Reception Section 131: Keyboard 132: Mouse 140: Display 150: Storage section 151: User Information 160: Control Unit 161: Input operation reception unit 162: Transceiver Unit 163: Notification Control Unit 164: Data Processing Unit 201: Communications Department 202: Storage section 203: Control Unit 901: Processor 902: Main memory 903 :Auxiliary storage device 921: Communications bus 2021: Target Database 2022: Estimated Model 2031: Receiver control module 2032: Transmit control module 2033: Target Data Acquisition Module 2034: Target Data Input Module 2035: Estimated data acquisition module 2036: Estimated Data Output Module 2037: Estimation Model Learning Module

Claims

1. A program to be executed on a computer equipped with a processor and memory, for estimating the state of a subject with respect to a predetermined disease, The memory stores a database containing data relating to the subject, The program is provided to the processor: The steps include obtaining the aforementioned target data, The process involves inputting the subject data into an estimation model, which is a trained model that has learned from past subject data and data indicating the state of a predetermined disease in the subject in the past, and outputting estimated data indicating the disease subtype and progression stage, which are the state of the subject with respect to the predetermined disease. The steps include: acquiring the estimated data output by the estimation model; The steps of outputting the acquired estimated data and then executing the following: The estimation model is composed of a probabilistic model that has learned from past data of the subject and data indicating the state of a predetermined disease in the subject in the past. A program that, in the step of outputting estimated data using the estimation model, uses the probabilistic model to introduce a latent time axis and reproduce the time series structure, thereby estimating pseudo-longitudinal time series data from cross-sectional subject data and imputing missing values.

2. The aforementioned program, further, The program according to claim 1, which causes the estimation model to perform the step of learning past subject data and data indicating the state of a predetermined disease in the subject in the past.

3. The aforementioned program, further, A step of generating a prompt that causes a large-scale language model to output data indicating an interpretation of either or both of the past subject data, or data indicating the status of a predetermined disease in the past subject; The steps of providing the generated prompt to the large-scale language model and receiving a response from the large-scale language model are performed. The program according to claim 2, wherein in the step of training the estimation model, the program is trained on the output received from the large-scale language model.

4. The program according to claim 1, wherein the estimation model is comprised of a graphical model in which disease subtypes and progression stages related to a predetermined disease are defined in advance.

5. The program according to claim 4, wherein the graphical model determines model coefficients based on learned past subject data and data indicating the state of a predetermined disease in the subject in the past.

6. The program according to claim 1, wherein the estimation model is an estimation model that outputs estimated data indicating the disease subtype and progression stage, which are the dementia status of the subject.

7. The program according to claim 1, wherein the subject data includes either or both of the subject's lifestyle data and / or the subject's clinical data at a medical institution.

8. An information processing device comprising a control unit and a memory, which estimates the state of a subject with respect to a predetermined disease, The memory stores a database containing data relating to the subject, The control unit, The steps include obtaining the aforementioned target data, The process involves inputting the subject data into an estimation model, which is a trained model that has learned from past subject data and data indicating the state of a predetermined disease in the subject in the past, and outputting estimated data indicating the disease subtype and progression stage, which are the state of the subject with respect to the predetermined disease. The steps include: acquiring the estimated data output by the estimation model; The steps of outputting the acquired estimated data and then performing the following steps are performed: The estimation model is composed of a probabilistic model that has learned from past data of the subject and data indicating the state of a predetermined disease in the subject in the past. An information processing device that, in the step of outputting estimated data using the estimation model, uses the probabilistic model to introduce a latent time axis and reproduce the time series structure, thereby estimating pseudo-longitudinal time series data from cross-sectional subject data and imputing missing values.

9. A method for estimating the state of a subject with respect to a predetermined disease, which is performed by a computer comprising a processor and memory, The memory stores a database containing data relating to the subject, The above method involves the processor, The steps include obtaining the aforementioned target data, The process involves inputting the subject data into an estimation model, which is a trained model that has learned from past subject data and data indicating the state of a predetermined disease in the subject in the past, and outputting estimated data indicating the disease subtype and progression stage, which are the state of the subject with respect to the predetermined disease. The steps include: acquiring the estimated data output by the estimation model; The steps of outputting the acquired estimated data and then performing the following steps are performed: The estimation model is composed of a probabilistic model that has learned from past data of the subject and data indicating the state of a predetermined disease in the subject in the past. A method for outputting estimated data using the estimation model, wherein, in the step of using the estimation model, a latent time axis is introduced using the probabilistic model to reproduce the time series structure, thereby estimating pseudo-longitudinal time series data from cross-sectional subject data and imputing missing values.

Images