System

Abstract

A system is provided.SOLUTION: The system includes an input means for inputting patient information, a means for generating a keyword on the basis of the inputted patient information, a retrieval means for acquiring case information from an external database by using the generated keyword, an analysis means for estimating the probability of a disease on the basis of the acquired case information, and a means for displaying the estimated probability information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In the medical field, vast amounts of information are being digitized, but it is difficult to quickly search for the necessary information and use it as a reference for diagnosis and treatment. In particular, there is a need for a system that can efficiently acquire case information, identify diseases and estimate their probability, and support the creation of medical certificates and prescriptions based on that information. To solve this problem, a method is needed that can search for information more quickly and accurately and support medical professionals. [Means for solving the problem]

[0005] The present invention is a system including the following means: an input means for inputting patient information; a means for generating keywords based on the input patient information; a search means for retrieving case information from an external database using the generated keywords; an analysis means for estimating disease probability based on the retrieved case information; and a means for displaying the estimated probability information. The system also includes a document generation means for automatically generating documents such as medical certificates, prescriptions, and referral letters; a means for displaying the generated documents in an editable format; and a learning means for training a machine learning model using stored data to improve the accuracy of the system. This improves the efficiency of disease probability estimation and document creation, thereby reducing the burden on medical professionals.

[0006] "Patient information" refers to information entered by medical professionals, such as basic patient information, symptoms, medical history, and test results.

[0007] "Input means" refers to an interface or device for obtaining patient information.

[0008] "Means for generating keywords" refers to a device or program that has the function of analyzing input patient information and generating related search keywords.

[0009] "Search means" refers to a device or program that searches for case information in an external database or on the Internet using the generated keywords.

[0010] "Case information" refers to past records and literature on related diseases and treatments.

[0011] "Analysis means" refers to a device or program that statistically estimates the probability of disease based on acquired case information.

[0012] "Display means" refers to a device or program for visually presenting analysis results and other information to a user.

[0013] "Document generation means" refers to a device or program that has the function of automatically generating documents such as medical certificates, prescriptions, and referral letters.

[0014] "Editable format" refers to a format that allows medical professionals to review the generated document and make corrections or additions as necessary.

[0015] "Learning means" refers to devices or programs that train machine learning models based on stored data and improve the accuracy of the system. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0018] First, the terms used in the following description will be explained.

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and assists them in estimating disease probability and creating documents based on that information. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means.

[0038] 1. Entering patient information and initial processing

[0039] User

[0040] The user (healthcare worker) logs in to a dedicated terminal (PC, tablet, etc.) and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, current test results, etc. Test results can be captured using a scanner or entered directly into the terminal.

[0041] Terminal

[0042] The device converts the input information into a standard format and sends it to the server, encrypting the data to prevent it from leaking, and notifying the user that the data has been sent.

[0043] 2. Data Acquisition and Analysis

[0044] server

[0045] The server stores the patient information sent from the device in a database. It also performs a preliminary data consistency check. It automatically generates keywords for related symptoms based on the patient information. This keyword generation is performed using natural language processing (NLP) to extract key keywords from symptoms and test results.

[0046] Based on the generated keywords, the server searches major literature databases and case report sites on the Internet, using medical database APIs and search engine queries. Related cases and literature are retrieved from the search results, and the results are filtered to extract the most relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[0047] 3. Probability estimation and information provision

[0048] server

[0049] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, "The probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%." These estimated results are converted into JSON format and sent to the terminal.

[0050] Terminal

[0051] The device visually displays the received information so that the user can check it. Possible display formats include lists and graphs. The device also automatically generates documents such as medical certificates, prescriptions, and referral letters based on the inference results. The generated documents are displayed in an editable format so that the user can check and modify them.

[0052] 4. Document Creation and Editing

[0053] server

[0054] The server prepares templates for medical certificates and prescriptions based on the inference results, and automatically generates draft documents by inserting the patient's individual information. This document generation is based on predefined templates.

[0055] Terminal

[0056] The terminal displays the generated draft document to the user and provides an interface that allows for corrections and additions as necessary. The user then checks and corrects the document and finalizes it based on the results.

[0057] 5. Data accumulation and accuracy improvement

[0058] server

[0059] The server stores the final diagnosis and treatment progress in a database. The stored data is then used to train the machine learning model. The accuracy of the system is improved by periodically retraining the machine learning model with new data sets.

[0060] As a specific example, if a patient complains of "headache, fever, cough," the user enters this information into the terminal and the information is sent to the server. The server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, this information is sent to the terminal for the user to confirm. Furthermore, the necessary medical certificates and prescriptions are automatically generated based on this information, and the user makes the final confirmation.

[0061] In this way, the system can provide information quickly and accurately, reducing the burden on medical professionals.

[0062] The processing flow will be explained below.

[0063] Step 1: Enter patient information

[0064] User

[0065] The user (medical worker) logs in to a dedicated terminal.

[0066] The user enters the patient's basic information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results.

[0067] Step 2: Sending data

[0068] Terminal

[0069] The terminal converts the entered patient information into a standard format.

[0070] The terminal transmits information to the server using encrypted communication.

[0071] The terminal notifies the user that the transmission is complete.

[0072] Step 3: Save data and check its integrity

[0073] server

[0074] The server stores the received patient information in a database.

[0075] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[0076] Step 4: Generate keywords

[0077] server

[0078] The server analyzes the stored patient information and generates keywords for relevant symptoms.

[0079] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[0080] Step 5: Internet Search

[0081] server

[0082] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[0083] Use medical database APIs or search engine queries to retrieve relevant case information.

[0084] Step 6: Filtering case information

[0085] server

[0086] The server performs filtering processing to extract highly relevant information from the acquired case information.

[0087] Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[0088] Step 7: Estimate the probability of disease

[0089] server

[0090] The server performs statistical analysis based on the filtered case information.

[0091] The probability of each disease is estimated, for example, "40% probability of influenza," "30% probability of cold," and "20% probability of pneumonia."

[0092] Step 8: Submit your information

[0093] server

[0094] The server converts the estimation results into JSON format and sends them to the terminal.

[0095] Step 9: Viewing Information

[0096] Terminal

[0097] The device visually displays the received information, for example in the form of a list or graph.

[0098] The user reviews and evaluates the displayed information.

[0099] Step 10: Prepare for document generation

[0100] server

[0101] The server prepares templates for medical certificates, prescriptions, referral letters, etc. based on the estimation results.

[0102] Insert the patient's individual information into the template and generate a draft document.

[0103] Step 11: Editing the Document

[0104] Terminal

[0105] The terminal displays the generated draft document to the user and provides it in an editable format.

[0106] The user checks the draft document and makes any necessary corrections.

[0107] Step 12: Save your data

[0108] server

[0109] The server stores the final diagnosis results and treatment progress in a database.

[0110] Step 13: Improve the accuracy of the system

[0111] server

[0112] The server uses the stored data to train the machine learning model.

[0113] The model is periodically retrained with new datasets to improve the accuracy of the system.

[0114] Example 1

[0115] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0116] In conventional medical information processing systems, medical professionals had to manually input patient information and then create diagnoses and documents based on that information, which was a cumbersome process. It also required a great deal of time and effort to efficiently search for related case information based on individual patient information and obtain appropriate diagnostic results. Furthermore, the system lacked the functionality to automatically generate and edit documents based on diagnostic results, increasing the burden on medical professionals. In addition, improving the accuracy of machine learning models also required manual effort, making the automation and accuracy improvement of the entire system a challenge.

[0117] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0118] In this invention, the server includes an information input means for inputting patient information, a keyword generation means for generating keywords based on the input patient information, a data acquisition means for acquiring case information from an external database using the generated keywords, a probability estimation means for estimating the probability of disease based on the acquired case information, an information display means for displaying the estimated probability information, a document generation means for automatically generating a medical certificate or prescription based on the estimation result, a document editing means for displaying the document generated by the document generation means in a format that can be edited by the user, and a learning means for improving the accuracy of the system using stored data. This reduces the burden on medical professionals and enables efficient and accurate diagnosis and document creation.

[0119] "Patient information" is a general term for a patient's basic personal information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[0120] "Information input means" refers to devices and software that allow a user to input patient information, and includes terminals and special input software.

[0121] "Keyword generation means" refers to a means for automatically extracting key keywords from related symptoms and test results based on input patient information.

[0122] "Data acquisition means" refers to a means for searching and acquiring case information from external databases or the Internet using the generated keywords.

[0123] "Probability estimation means" refers to a means of estimating the probability of disease occurrence based on acquired case information using statistical analysis and machine learning techniques.

[0124] "Information display means" refers to a device or software for visually displaying estimated probability information to a user, including display in list or graph format.

[0125] "Document generation means" refers to software or algorithms for automatically generating medical certificates and prescriptions based on the inference results.

[0126] The "document editing means" refers to a means for providing an interface that allows a user to modify or add to a generated document.

[0127] "Learning means" refers to the means of training a machine learning model using stored patient information and diagnostic results to improve the accuracy of the system.

[0128] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and, based on that information, assists in disease probability estimation and document creation. The system includes an information input means for inputting patient information, a keyword generation means for generating keywords, a data acquisition means for searching case information, a probability estimation means for analyzing the acquired information, an information display means for displaying the analysis results, a document generation means for automatically generating documents, a document editing means for editing the generated documents, and a learning means for improving the accuracy of the system.

[0129] Hardware and Software Configuration

[0130] Information input method

[0131] User:

[0132] Users (healthcare workers) log in to a dedicated terminal (PC, tablet, etc.), which is equipped with an easy-to-use GUI (graphical user interface), and enter basic patient information (name, age, gender, etc.) and specific symptoms (e.g., headache, fever, cough).

[0133] Device:

[0134] The device has a software module that converts the information into a standard format, such as JSON or XML, and then the converted data is encrypted and securely sent to the server.

[0135] Probability estimation means and information display means

[0136] server:

[0137] The server stores the received patient information in a database and performs an initial consistency check. It then uses natural language processing (NLP) technology to extract relevant keywords from the data. The generated keywords are then used to search for related cases and literature using medical database APIs and search engines. After filtering, the most reliable information is used to estimate the probability of disease using statistical analysis and machine learning algorithms.

[0138] Device:

[0139] The terminal visually displays the estimation results, which can be displayed in list or graph format for the user to review.

[0140] Document creation and editing tools

[0141] server:

[0142] The server prepares templates for medical certificates and prescriptions based on the estimation results, and automatically generates draft documents by inserting the patient's individual information. These templates are predefined.

[0143] Device:

[0144] The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections or additions. The user can then check, correct, and finalize the document.

[0145] Learning tools

[0146] server:

[0147] The server uses the stored data to train the machine learning model, which is periodically retrained with new data sets to improve the system's accuracy. This retraining process uses advanced machine learning techniques such as deep learning.

[0148] Specific examples

[0149] For example, if a patient complains of "headache, fever, cough," the following process occurs:

[0150] User:

[0151] Symptoms are entered into the terminal, which converts them into a standard format and sends them to the server via encrypted communication.

[0152] server:

[0153] It receives information, generates keywords such as "headache," "fever," and "cough," and searches and filters related cases and literature. For example, it estimates the probability of influenza at 40%, cold at 30%, and pneumonia at 20%.

[0154] Device:

[0155] The system displays the estimated results and automatically generates the necessary medical certificates and prescriptions, which the user can then review and finalize.

[0156] Prompt Sentence Examples

[0157] "The patient's symptoms are headache, fever, and cough. Past medical history includes allergic rhinitis, and current test results show an elevated white blood cell count. Based on this information, generate search keywords and related case information, and calculate the estimated probability of the most relevant disease. Additionally, create a draft medical certificate based on the results."

[0158] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0159] Step 1:

[0160] The user enters information

[0161] Input: Patient basic information (name, age, gender, etc.), specific symptoms (e.g. headache, fever, cough), past medical history, and current test results.

[0162] How it works: The user (healthcare worker) logs in to a dedicated device (PC, tablet, etc.) and enters patient information. The user enters information using an easy-to-use GUI.

[0163] Output: The entered patient information is temporarily saved in the terminal's memory.

[0164] Step 2:

[0165] The device prepares the data

[0166] Input: Patient information entered by the user.

[0167] How it works: The device converts the input information into a standard format (e.g., JSON or XML) to facilitate further processing. The converted data is then encrypted and security protocols are applied to prevent external access.

[0168] Output: The encrypted data converted into a standard format.

[0169] Step 3:

[0170] The device sends the data to the server

[0171] Input: Encrypted data converted to a standard format.

[0172] Operation: The device sends encrypted data to the server. The communication protocol is SSL / TLS, and encryption is used to prevent unauthorized access or tampering of data. After the transmission is complete, the device notifies the user that the transmission was successful.

[0173] Output: Data sent to the server and notification to the user that the data has been sent.

[0174] Step 4:

[0175] The server receives and stores the data

[0176] Input: Encrypted data sent from the terminal.

[0177] Operation: The server decrypts the received data and stores it in the database. A relational database is used as the database system. Before storing it in the database, the data is checked for consistency, and an error log is recorded if there is an inconsistency.

[0178] Output: Patient information stored in database and error log (if required).

[0179] Step 5:

[0180] The server generates the keywords

[0181] Input: Patient information stored in the database.

[0182] How it works: The server uses natural language processing (NLP) techniques to extract key keywords from the patient information entered. For example, if the patient information contains "headache," "fever," and "cough," those keywords are generated.

[0183] Output: The generated keywords.

[0184] Step 6:

[0185] The server searches for case information

[0186] Input: Generated keywords.

[0187] How it works: The server uses the generated keywords to search for case reports and literature on the Internet using medical database APIs and search engines. It queries specific databases (e.g., medical research databases or academic paper databases) to obtain relevant information. The obtained data is then filtered to select information that is highly reliable and closely matches the keywords.

[0188] Output: Filtered case information.

[0189] Step 7:

[0190] The server estimates the probability of the disease

[0191] Input: Filtered case information.

[0192] How it works: The server uses statistical analysis and machine learning algorithms to estimate disease probabilities based on filtered case information, such as "the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%."

[0193] Output: Estimated probability information.

[0194] Step 8:

[0195] The server sends the estimation results to the device.

[0196] Input: Estimated probability information.

[0197] How it works: The server converts the probability information into JSON format and sends it to the device. The transmission process is secure using the SSL / TLS protocol.

[0198] Output: Probability information sent to the terminal.

[0199] Step 9:

[0200] The device displays the estimated results.

[0201] Input: Probability information sent to the terminal.

[0202] Operation: The device visually displays the received probability information in a list or graph format designed to be easily understood by the user.

[0203] Output: Probability information displayed to the user.

[0204] Step 10:

[0205] The server automatically generates the document

[0206] Input: Estimated probability information and basic patient information.

[0207] How it works: The server prepares templates for medical certificates and prescriptions based on the inference results, embeds the patient's individual information, and automatically generates draft documents using a template engine.

[0208] Output: Auto-generated draft document.

[0209] Step 11:

[0210] The device edits and displays the document.

[0211] Input: Auto-generated draft document.

[0212] Operation: The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections and additions. Once editing is complete, the user performs a final check and finalizes the document.

[0213] Output: Final revised and verified document.

[0214] Step 12:

[0215] The server stores the data and updates the learning model.

[0216] Input: Finalized documentation and diagnostic results.

[0217] How it works: The server stores the final diagnosis and treatment progress in a database. The stored data is used to periodically retrain the machine learning model to improve the accuracy of the system. The machine learning algorithm includes adaptive learning to accommodate increasing data.

[0218] Output: Updated machine learning model and saved diagnostic data.

[0219] (Application example 1)

[0220] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0221] In modern factories, maintenance and inspection of each machine is still often done manually. This relies on the knowledge and experience of the maintenance personnel, which can delay early detection of abnormalities and optimal maintenance responses. Furthermore, updating maintenance histories and manuals is time-consuming, which tends to lower the efficiency of the entire factory. Furthermore, identifying the cause of an abnormality and selecting a countermeasure can take time, so an efficient system to solve these problems is needed.

[0222] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0223] In this invention, the server includes input means for inputting patient information, means for generating keywords based on the input patient information, search means for retrieving case information from an external database using the generated keywords, analysis means for estimating the probability of disease based on the retrieved case information, means for displaying the estimated probability information, input means for inputting each device information, means for generating keywords based on the input device information, search means for retrieving maintenance response actions from an external database using the generated keywords, and means for applying the retrieved maintenance response actions to machine maintenance and repair. This makes it possible to efficiently and quickly identify machine abnormalities and propose optimal maintenance responses.

[0224] "Patient information" refers to basic patient data (such as name, age, sex, past medical history, specific symptoms, and test results) entered by medical professionals.

[0225] "Input means" refers to equipment or devices (PCs, tablets, smartphones, etc.) used to collect patient information and information on machinery and equipment and input it as data.

[0226] "Keyword generation means" refers to the technology or algorithm used to extract key keywords from input information.

[0227] A "search means" is a system or software that uses the generated keywords to retrieve related information from an external database.

[0228] "Analysis means" refers to a device or program that has the function of estimating and displaying specific results using statistical analysis or machine learning based on acquired information.

[0229] "Display means" refers to a display or monitor for visually displaying analysis results, estimated probability information, maintenance response actions, etc.

[0230] "Each device information" refers to data on the status and abnormalities of various machinery and equipment within the factory (such as years of operation, error logs, and current operating status).

[0231] "Maintenance response actions" refer to specific procedures and tasks for resolving abnormalities or breakdowns in machinery and equipment.

[0232] "Document generation means" refers to software or a system for automatically generating documents such as medical certificates, prescriptions, and maintenance manuals based on acquired information.

[0233] "Learning" is the process of using stored data to train machine learning models to improve the system's analytical accuracy and predictive capabilities.

[0234] This invention relates to a "smart maintenance system" for improving the efficiency of maintenance and inspection work for factory robots and machinery equipment. In particular, it is a system for automating a series of processes including input of equipment information, keyword generation, related information search, analysis, display, document generation, and learning.

[0235] 1. Entering device information and initial processing

[0236] User

[0237] Users (maintenance personnel) use dedicated devices (smartphones, tablets, etc.) to input information about each piece of machinery and equipment. For example, they input data such as the machine's name, years of operation, current status, and error logs. This information plays an important role in subsequent analysis and maintenance measures.

[0238] Terminal

[0239] The terminal converts the input data into a standard format and transmits it to the server using secure encrypted communications. After transmitting the data, the terminal notifies the user that the transmission has been completed.

[0240] 2. Data Acquisition and Analysis

[0241] server

[0242] The server stores the machine equipment data sent from the terminal in a database. At the same time, it checks the data for consistency. Next, it uses natural language processing (NLP) to generate keywords from the input information. For example, it extracts key keywords from error logs, such as "loud operating noise" or "vibration felt."

[0243] Based on the generated keywords, the server searches an external database for relevant maintenance actions, filtering the results based on criteria such as reliable data sources and keyword match.

[0244] 3. Estimation and provision of information on maintenance response actions

[0245] server

[0246] The server analyzes the filtered information and predicts the most appropriate maintenance action. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." The predicted maintenance action is sent to the terminal in JSON format.

[0247] Terminal

[0248] The terminal visually displays the received maintenance response actions. Possible display formats include lists, graphs, and text. In addition, documents such as maintenance manuals and maintenance histories are automatically generated based on the estimation results. The generated documents are displayed in a format that can be checked and edited by the user.

[0249] 4. Document Creation and Editing

[0250] server

[0251] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This document generation is performed based on predefined templates.

[0252] Terminal

[0253] The terminal displays the generated draft document to the user and provides an editing interface that allows the user to make corrections or additions as necessary. After the user checks and corrects the document, it is finalized.

[0254] 5. Data accumulation and accuracy improvement

[0255] server

[0256] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved.

[0257] Specific examples

[0258] The maintenance technician enters the status information of the "Robot Arm A1" (2 years of operation, operating noise, vibration detection) into the terminal. The input data is converted to JSON format and sent to the maintenance server.

[0259] The server analyzes the error log information sent and extracts key keywords ("operating sound" and "vibration") using TF-IDF. Maintenance response actions are searched and filtered based on the obtained keywords.

[0260] Based on the extracted keywords, the most appropriate maintenance actions ("Check the cooling system and clean the filters" and "Check for loose parts and lubricate the bearings") are displayed.

[0261] Prompt Sentence Examples

[0262] "You input the machine's condition, for example, 'it's loud when it's running' or 'I feel vibrations.' Then, based on this information, we suggest the best maintenance action."

[0263] As described above, the present invention provides a system that improves the efficiency of maintenance management within a factory, reduces the burden on maintenance personnel, and enables early detection of machine abnormalities.

[0264] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0265] Step 1:

[0266] Users input information about each piece of machinery and equipment using a dedicated device (smartphone, tablet, etc.). The input includes the machine's name, years of operation, current status, error log, etc. This allows data input (machine information) from the user to be obtained. This data is the basic data required for subsequent analysis.

[0267] Step 2:

[0268] The terminal converts the machine information entered by the user into a standard format, for example, JSON format, and sends it to the server using encrypted communication. This allows data to be sent from the terminal to the server.

[0269] Step 3:

[0270] The server receives the machine information sent from the terminal and stores it in a database. Before storing it, it checks the data for consistency and ensures that there are no abnormalities. This ensures the reliability of the data.

[0271] Step 4:

[0272] The server performs natural language processing (NLP) based on the received machine information to extract key keywords. For example, keywords such as "loud operating noise" and "vibration felt" are generated from error logs using TF-IDF. This provides basic keywords for analysis.

[0273] Step 5:

[0274] The server queries an external database (such as a maintenance manual database) based on the generated keywords. It extracts appropriate maintenance response actions and collects information on them. Search results are filtered based on reliable data sources, keyword matching, etc.

[0275] Step 6:

[0276] The server performs statistical analysis on the acquired maintenance response actions and estimates specific maintenance measures. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." This allows it to estimate the optimal maintenance response action.

[0277] Step 7:

[0278] The server converts the estimated maintenance action into JSON format and sends it to the terminal, which then transmits the analysis results to the terminal.

[0279] Step 8:

[0280] The terminal visually displays the received maintenance action and notifies the user. The display format can be a list, graph, text, etc. This allows the user to confirm the maintenance action.

[0281] Step 9:

[0282] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This generates documents for maintenance actions.

[0283] Step 10:

[0284] The terminal displays the generated draft document to the user and provides an editing interface that allows corrections and additions as necessary. After the user checks and corrects the document, they finalize it. This finalizes the maintenance document.

[0285] Step 11:

[0286] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved, thereby improving overall system performance.

[0287] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0288] The present invention relates to an information provision system that efficiently inputs patient information and, based on that information, estimates disease probability, creates documents, and recognizes user emotions. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means. Furthermore, it incorporates an emotion engine and has the function of customizing the method of providing diagnosis results and advice based on the user's emotional state.

[0289] 1. Entering patient information and initial processing

[0290] User

[0291] The user (healthcare worker) logs in to a dedicated terminal and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly into the terminal.

[0292] Terminal

[0293] The terminal converts the input information into a standard format and transmits it to the server using encrypted communications. Once transmission is complete, the terminal notifies the user.

[0294] 2. Data Acquisition and Analysis

[0295] server

[0296] The server stores the patient information sent from the terminal in a database, checks the data consistency, and sends a supplemental request to the terminal if any information is missing.

[0297] server

[0298] The server analyzes stored patient information and generates keywords for related symptoms, using natural language processing (NLP) to extract key keywords from symptoms and test results.

[0299] server

[0300] The server searches online literature databases and case report sites based on the generated keywords, and retrieves relevant case information using medical database APIs and search engine queries.

[0301] server

[0302] The server performs filtering to extract highly relevant information from the acquired case information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[0303] server

[0304] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, it may estimate "40% probability of influenza," "30% probability of cold," or "20% probability of pneumonia." These estimation results are converted into JSON format and sent to the device.

[0305] 3. Information provision and document generation

[0306] Terminal

[0307] The device visually displays the received information so that the user can check it. It can use display formats such as lists and graphs. Furthermore, it automatically generates medical certificates, prescriptions, referral letters, etc. based on the inference results.

[0308] server

[0309] Based on the estimation results, the server prepares templates for medical certificates and prescriptions, inserts the patient's individual information, and generates draft documents.

[0310] Terminal

[0311] The terminal provides the generated draft document in an editable format, allowing the user to check and modify the draft document and make any necessary additions.

[0312] 4. Emotion Recognition and Response

[0313] Terminal

[0314] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[0315] server

[0316] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[0317] server

[0318] Based on the assessment, the server adjusts the presentation of the diagnostic results to provide information in a format that is most acceptable to the user, for example, using calmer, more reassuring language for a user in a stressed state.

[0319] Terminal

[0320] The device uses an emotion engine to provide users with customized medical advice, for example, providing detailed instructions on specific steps to take or the need for additional testing if the user is worried.

[0321] 5. Data accumulation and learning

[0322] server

[0323] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[0324] As a concrete example, if a patient complains of "headache, fever, and cough," the user enters this information into the terminal, and the server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, the terminal will display this information and further analyze the user's emotional state to provide information in the most optimal form. Medical certificates and prescriptions are also automatically generated, with the user making the final confirmation.

[0325] This system can provide information quickly and accurately, reducing the burden on medical professionals. Furthermore, emotion recognition can be used to present information in a way that is easy for users to accept, improving the quality of medical care.

[0326] The processing flow will be explained below.

[0327] Step 1: Enter patient information

[0328] User

[0329] The user (medical worker) logs in to a dedicated terminal.

[0330] The user enters the patient's basic information (name, age, gender, etc.), symptoms (headache, fever, cough, etc.), past medical history, and current test results.

[0331] Step 2: Sending data

[0332] Terminal

[0333] The terminal converts the entered patient information into a standard format.

[0334] The terminal transmits the patient information to the server using encrypted communication.

[0335] The terminal notifies the user that the transmission is complete.

[0336] Step 3: Save data and check its integrity

[0337] server

[0338] The server stores the received patient information in a database.

[0339] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[0340] Step 4: Generate keywords

[0341] server

[0342] The server analyzes stored patient information and automatically generates keywords for relevant symptoms.

[0343] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[0344] Step 5: Internet Search

[0345] server

[0346] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[0347] Use medical database APIs or search engine queries to retrieve relevant case information.

[0348] Step 6: Filtering case information

[0349] server

[0350] The server performs filtering to extract highly relevant information from the acquired case information.

[0351] Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[0352] Step 7: Estimate the probability of disease

[0353] server

[0354] The server performs statistical analysis based on the filtered case information and estimates the probability of each disease.

[0355] For example, estimates might say, "The probability of influenza is 40%," "The probability of a cold is 30%," and "The probability of pneumonia is 20%."

[0356] Step 8: Submit your information

[0357] server

[0358] The server converts the estimated probability information into JSON format and sends it to the terminal.

[0359] Step 9: Viewing Information

[0360] Terminal

[0361] The device visually displays the information it receives, for example in the form of a list or graph.

[0362] The user reviews and evaluates the displayed information.

[0363] Step 10: Emotion Recognition

[0364] Terminal

[0365] The device uses an emotion engine to analyze the user's facial expressions, tone of voice, and input content while operating the device, and recognizes the user's emotional state.

[0366] Step 11: Assess and regulate your emotional state

[0367] server

[0368] The server evaluates the user's emotional state (stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[0369] The server adjusts the way the information is presented based on the evaluation results, and provides the information in a format that is easy for the user to accept.

[0370] Step 12: Customize your information offering

[0371] Terminal

[0372] The terminal uses the emotion engine to present customized medical advice to the user.

[0373] For example, a user in a good emotional state may be provided with concise and efficient information, while a user experiencing stress may be provided with detailed and reassuring information.

[0374] Step 13: Document Generation

[0375] server

[0376] The server prepares templates for a diagnosis, prescription, and referral letter based on the estimation results and the user's emotional state.

[0377] Insert the patient's individual information into the template and automatically generate a draft document.

[0378] Step 14: Edit and review the document

[0379] Terminal

[0380] The terminal displays the generated draft document to the user and provides it in an editable format.

[0381] The user checks and corrects the draft document and makes any necessary additions.

[0382] Step 15: Save your data

[0383] server

[0384] The server stores the final diagnosis results and treatment progress in a database.

[0385] Step 16: Training the system and improving accuracy

[0386] server

[0387] The server uses the stored data to train the machine learning model.

[0388] The model is periodically retrained with new datasets to improve the accuracy of the system.

[0389] Example 2

[0390] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0391] In modern medical settings, entering and analyzing patient information, estimating diagnostic probabilities, and determining optimal treatment plans requires a great deal of time and effort. Furthermore, there is often a lack of information provided that takes into account the patient's emotions and mental state, which increases the burden on medical professionals and increases patient anxiety. To efficiently resolve these issues and improve the quality of medical care, a comprehensive and flexible information provision system is required.

[0392] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0393] In this invention, the server includes means for inputting patient information, means for converting the input patient information into a standard format and transmitting it, means for saving the transmitted patient information, means for generating keywords based on the saved patient information, means for retrieving case information from an external database using the generated keywords, means for filtering the retrieved case information based on relevance, means for estimating disease probability based on the filtered case information, means for displaying the estimated probability information, and means for analyzing user emotions and adjusting the information presentation method, thereby enabling efficient input and analysis of patient information, accurate disease probability estimation, and information provision that takes patient emotions into consideration.

[0394] "Patient information" is a general term for basic patient information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[0395] "Means for inputting" refers to a device or interface that allows a user to input patient information.

[0396] "Means for converting into a standard format and transmitting" refers to the function for converting input patient information into a specified format and transmitting it securely to a server using an encrypted communication protocol.

[0397] "Means for storing" refers to a function for storing transmitted patient information in a database.

[0398] "Means for generating keywords" refers to technology that analyzes stored patient information and extracts keywords based on key symptoms and test results.

[0399] "Means for obtaining case information" refers to a function for obtaining related case information from external databases or case report sites using the generated keywords.

[0400] "Filtering means" refers to a function for sorting acquired case information based on relevance and extracting only highly reliable information.

[0401] "Means for estimating the probability of disease" refers to technology for performing statistical analysis based on filtered case information to calculate the probability of occurrence of a specific disease.

[0402] "Means for displaying probability information" refers to a display interface that allows a user to visually confirm the estimated disease occurrence probability.

[0403] "Means for analyzing emotions and adjusting the way information is presented" refers to a function that analyzes the user's emotional state, such as facial expressions and tone of voice, and customizes the way information is presented based on the results.

[0404] "Means for automatically generating documents" refers to a function for automatically generating medical documents such as medical certificates, prescriptions, and referral letters.

[0405] "Means for displaying in an editable format" refers to an interface for displaying the generated medical document in a format that can be checked and corrected by the user.

[0406] "Means for training machine learning models to improve the accuracy of the system" refers to technology that uses stored data on past diagnostic results and treatment progress to train machine learning models and improve the diagnostic accuracy of the system.

[0407] The present invention relates to an information providing system that efficiently inputs patient information and, based on the input information, estimates the probability of disease, creates documents, and recognizes user emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[0408] System configuration and basic operations

[0409] User

[0410] The user (healthcare worker) logs in to a dedicated terminal. The terminal provides an interface for entering basic patient information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly by hand. Once the user has finished entering the patient information, the terminal converts this information into a standard format and sends it to the server using an encrypted communication protocol (e.g., HTTPS).

[0411] Terminal

[0412] The terminal converts the received patient information into a standard JSON format and sends the information to the server using the HTTPS protocol. Once the transmission is complete, a notification of the completion is displayed to the user.

[0413] server

[0414] The server stores the patient information sent from the device in a database. It checks the consistency of the stored information and sends a supplemental request to the device if any information is missing. After consistency is confirmed, the server uses natural language processing (NLP) technology (e.g., spaCy or NLTK) to generate keywords from the patient information based on key symptoms and test results. Using the generated keywords, the server retrieves relevant case information from external literature databases and case report sites via medical database APIs (e.g., PubMed, ClinicalTrials.gov) or search engines.

[0415] The server filters the acquired case information and selects only the most relevant information. Based on the filtered information, the server performs statistical analysis to estimate the probability of a specific disease occurring. The estimated results (e.g., "40% probability of influenza," "30% probability of cold," "20% probability of pneumonia") are converted to JSON format and sent back to the terminal.

[0416] Linking devices and servers

[0417] The terminal visually displays the received probability estimation results, using formats such as lists and graphs. Furthermore, based on the information sent from the server, the terminal automatically generates medical documents such as medical certificates, prescriptions, and referral letters. The generated documents are provided in an editable format that the user can review and modify.

[0418] Emotion recognition function

[0419] When the user operates the device, the device uses an emotion engine to recognize the user's emotional state, such as facial expressions and tone of voice. The server evaluates the user's emotional state based on the analysis results of the emotion engine and adjusts the way the diagnosis results are presented. For example, if the user is anxious, the system will display information in calm language and provide detailed instructions on how to deal with the situation.

[0420] Learning Function

[0421] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model (e.g., neural network) to improve the accuracy of the system.

[0422] Specific examples

[0423] For example, if a patient complains of headache, fever, and cough, the user enters this information into the device. The server uses natural language processing to automatically generate keywords such as "headache," "fever," and "cough," and searches the case database. The search results include an estimate of a 40% probability of influenza, a 30% probability of a cold, and a 20% probability of pneumonia. These estimates are sent to the device in JSON format, where they are displayed visually in graphs. The emotion engine analyzes the user's facial expressions and provides customized advice, such as displaying the diagnosis results in calm language. Furthermore, a medical certificate and prescription are automatically generated, and the user makes the final confirmation.

[0424] Example prompts for generative AI models

[0425] "A patient complains of headache, fever, and cough. Their only past medical history is allergic rhinitis. Estimate the probability of influenza and visually display the results. Also, if the user is stressed, please respond by providing information in a calmer tone."

[0426] This system enables prompt and accurate information provision, reducing the burden on medical professionals. In addition, the emotion recognition function allows information to be provided in a format that is easy for users to accept, improving the quality of medical care.

[0427] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0428] Step 1:

[0429] User login and patient information entry

[0430] User

[0431] The user logs in to a dedicated terminal. They access the login screen, enter their user ID and password, and click the login button. If login is successful, they are redirected to the patient information input screen. The user enters the patient's basic information (name, age, sex), specific symptoms (headache, fever, cough, etc.), past medical history, and current test results in the form on the screen, and clicks the "Submit" button.

[0432] Input and Output

[0433] Input: Patient information entered by the user into the terminal

[0434] Output: The input information is converted into a standard format.

[0435] Step 2:

[0436] Converting and sending input information

[0437] Terminal

[0438] The terminal converts the received patient information into a standard format (e.g., JSON format) and sends it to the server using an encrypted communication protocol (e.g., HTTPS). Once the transmission is complete, the terminal displays a pop-up message stating, "Information transmission completed."

[0439] Input and Output

[0440] Input: User-entered patient information

[0441] Output: Information converted to standard format and notification of completion of transmission

[0442] Step 3:

[0443] Server information storage and supplementary requests

[0444] server

[0445] The server receives the patient information sent from the terminal and stores it in the database. After storing it, it checks the consistency of the data and sends a supplementary request to the terminal if necessary information is missing.

[0446] Input and Output

[0447] Input: Patient information submitted in a standard format

[0448] Output: Stored information and supplementary requests after consistency check

[0449] Step 4:

[0450] Generate keywords for patient information

[0451] server

[0452] The server analyzes the stored patient information and generates keywords based on relevant symptoms and test results using natural language processing (NLP) techniques (e.g., spaCy and NLTK).

[0453] Input and Output

[0454] Input: Saved patient information

[0455] Output: Generated keyword list

[0456] Step 5:

[0457] Case database search

[0458] server

[0459] The server searches external literature databases and case report sites based on the generated keywords, using medical database APIs and search engine queries.

[0460] Input and Output

[0461] Input: Generated keyword list

[0462] Output: Retrieved case information

[0463] Step 6:

[0464] Case information filtering

[0465] server

[0466] The server then performs a filtering process based on the acquired case information to extract highly relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[0467] Input and Output

[0468] Input: Retrieved case information

[0469] Output: Filtered case information list

[0470] Step 7:

[0471] Disease probability estimation and result transmission

[0472] server

[0473] The server performs statistical analysis based on the filtered case information to estimate the probability of candidate diseases. The estimated results are converted into JSON format and sent to the device.

[0474] Input and Output

[0475] Input: Filtered case information

[0476] Output: Estimated disease probability information

[0477] Step 8:

[0478] Displaying diagnostic results and generating documentation

[0479] Terminal

[0480] The device visually displays the received information for the user to review. Display formats include lists and graphs. Furthermore, medical documents such as medical certificates, prescriptions, and referral letters are automatically generated based on the inference results. The generated documents are provided in an editable format, allowing the user to review and correct them.

[0481] Input and Output

[0482] Input: Estimated disease probability information

[0483] Output: Visually displayed information and generated medical documents

[0484] Step 9:

[0485] Emotion recognition and information presentation adjustment

[0486] Terminal

[0487] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[0488] server

[0489] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine. Based on the evaluation results, it adjusts the way the diagnosis results are presented to provide information in a format that is most acceptable to the user. For example, for a user in a stressed state, it uses calmer, more reassuring language.

[0490] Terminal

[0491] The device uses an emotion engine to provide users with customized medical advice, detailing specific steps to take or the need for additional testing if the user is concerned.

[0492] Input and Output

[0493] Input: User's emotional state

[0494] Output: customized diagnostic results

[0495] Step 10:

[0496] Data accumulation and learning

[0497] server

[0498] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[0499] Input and Output

[0500] Input: Final diagnosis and treatment progress

[0501] Output: Updated machine learning model

[0502] Specific actions

[0503] In this way, through a series of processes, patient information is efficiently input and analyzed, appropriate disease probability is estimated, and optimal treatment plans are determined.In addition, by presenting information that takes the user's emotions into consideration, the burden on medical professionals is reduced and patients feel more at ease.

[0504] (Application example 2)

[0505] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0506] Conventional health management systems for factory workers have difficulty efficiently integrating multiple health data sets and quickly and accurately assessing health risks. Furthermore, they do not provide appropriate advice based on the worker's emotional state, resulting in decreased productivity and the risk of workers' health deteriorating. These issues needed to be resolved.

[0507] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0508] In this invention, the server includes an input means for inputting the health data of workers, a means for generating keywords based on the input health data, a search means for retrieving case information from an external database using the generated keywords, an analysis means for estimating the probability of a health risk based on the retrieved case information, a means for displaying the estimated probability information, an emotion recognition means for recognizing the emotional state of the user, a means for customizing the information provision method based on the recognized emotional state, a document generation means for automatically generating documents such as medical certificates and prescriptions, a means for displaying the generated documents in an editable format, and a means for estimating health risks based on the health data and suggesting preventive measures. This makes it possible to efficiently monitor and evaluate the health status of factory workers and provide appropriate advice according to their emotional state.

[0509] "Patient information" refers to basic information about an individual patient collected by a medical professional, including, specifically, data such as the patient's name, age, gender, symptoms, medical history, and current test results.

[0510] "Input means" refers to an interface or device that allows a user to input patient information into a terminal.

[0511] The "keyword generating means" is a means for extracting major symptoms and related information from input patient information and generating search keywords.

[0512] The "search means" is a means for retrieving related case information from an external database using the generated keywords.

[0513] The "analysis means" is a means for estimating the probability of a particular disease or health risk based on the acquired case information.

[0514] The "display means" is a means for visually displaying estimated probability information and diagnostic results in a format that can be confirmed by the user.

[0515] The "emotion recognition means" is a means for analyzing the user's facial expressions and tone of voice to recognize the user's emotional state.

[0516] The "information presentation method customization means" is a means for optimizing the information presentation method based on the recognized emotional state.

[0517] The "document generation means" is a means for automatically creating documents such as medical certificates, prescriptions, and referral letters.

[0518] "Means for displaying in an editable format" refers to means for providing the generated document in a format that allows the user to check and modify it.

[0519] "Healthcare data" refers to data that indicates the health status of factory workers, specifically including blood pressure, heart rate, symptoms, medical history, etc.

[0520] "Means for estimating health risks" are means for assessing the need for preventive measures and countermeasures based on health data and for calculating the probability of risk.

[0521] A "means for proposing preventive measures" is a means for proposing specific preventive measures or countermeasures based on estimated health risks.

[0522] This invention is applied to a health management system for factory workers. Its basic components include an input means for inputting health data, an analysis means for standardizing and analyzing the data, a display means for displaying and editing estimation results and documents, and an emotion recognition means for recognizing emotional states. It also includes a means for assessing health risks and suggesting individual preventive measures.

[0523] 1. Health data entry and initial processing

[0524] User

[0525] Users (factory workers or managers) use dedicated input devices (e.g., PCs or tablets) to enter health data into the system. This data includes blood pressure, heart rate, symptoms, medical history, etc. After completing the data entry, the user sends the data from their device to the server. At this time, the data is converted into a standard format and sent to the server using encrypted communication.

[0526] 2. Data Acquisition and Analysis

[0527] server

[0528] The server receives the health data sent from the device and stores it in a database. Natural language processing (NLP) is used to generate keywords from the stored data, and relevant case information is searched for in external databases. Search methods include medical database APIs and search engine queries.

[0529] server

[0530] The server filters the acquired case information and estimates health risks based on reliable information. For example, it may estimate "40% probability of overwork," "30% probability of high blood pressure," or "20% probability of myocardial infarction." The estimated results are converted into JSON format and sent to the device.

[0531] 3. Information provision and document generation

[0532] Terminal

[0533] The device visually displays the received information for the user to review, and can use lists and graphs as display formats. It can also automatically generate and display documents such as medical certificates, prescriptions, and referral letters in user-editable formats.

[0534] 4. Emotion Recognition and Response

[0535] Terminal

[0536] When the user operates the terminal, the terminal uses emotion recognition means to recognize the user's emotional state from the user's facial expression, tone of voice, input contents, etc.

[0537] server

[0538] Based on the analysis results of the emotion recognition means, the server evaluates the user's emotional state (e.g., stress, relief, fatigue, etc.) and optimizes the way it provides information. For example, if the user is fatigued, it will provide detailed, specific instructions on how to deal with the situation in a gentle manner.

[0539] Terminal

[0540] The device uses emotion recognition to provide customized advice to the user, detailing specific actions to take or the need for further testing if the user is concerned.

[0541] 5. Data accumulation and learning

[0542] server

[0543] The server stores the final diagnosis results and countermeasure results in a database and uses the stored data to train the machine learning model, thereby improving the accuracy of the system.

[0544] Examples:

[0545] For example, suppose Mr. Tanaka (not his real name) complains of fatigue, headache, and muscle pain during a health checkup. This information is entered into the terminal. Based on this, the server stores the keywords "fatigue," "headache," and "muscle pain," and retrieves case information from an external database. As a result, it estimates that there is a "40% chance of overwork," a "30% chance of high blood pressure," and a "20% chance of myocardial infarction," and Mr. Tanaka's emotional state is also analyzed. If Mr. Tanaka is feeling stressed, it gently explains how to deal with the situation.

[0546] Example prompt sentence:

[0547] "Design a system to monitor and analyze the health status of factory workers. Analyze the entered health data (e.g., symptoms, medical history, current test results) and estimate health risks. Additionally, analyze the user's emotional state and provide individually customized advice."

[0548] In this way, by using this system, it is possible to efficiently manage the health of factory workers and provide appropriate advice according to the emotional state of each individual worker.

[0549] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0550] Step 1:

[0551] The user enters health data.

[0552] Input: The user inputs health data such as blood pressure, heart rate, symptoms, and medical history into the device.

[0553] Processing: The entered data is converted into a standard format and sent to the server using encrypted communications.

[0554] Output: A confirmation of receipt of the data sent is displayed on the terminal.

[0555] Step 2:

[0556] The server receives and stores the data.

[0557] Input: Health data sent from the device in a standard format.

[0558] Processing: Stores the received data in a database, checks the data consistency, and sends a supplemental request to the terminal if there is missing information.

[0559] Output: The saved data. If additional information is required, the request is notified to the terminal.

[0560] Step 3:

[0561] The server generates the keywords.

[0562] Input: Stored health data.

[0563] Processing: Using natural language processing (NLP), key keywords are automatically extracted. For example, "fatigue," "headache," and "muscle pain" are extracted.

[0564] Output: The generated keywords.

[0565] Step 4:

[0566] The server retrieves case information from an external database.

[0567] Input: Generated keywords.

[0568] Processing: Search for case information based on keywords using medical database APIs and search engine queries.

[0569] Output: Retrieved case information.

[0570] Step 5:

[0571] The server estimates the health risk.

[0572] Input: Retrieved case information.

[0573] Processing: Filtering is performed and statistical analysis is carried out based on highly reliable information. For example, estimates such as "40% probability of overwork," "30% probability of high blood pressure," and "20% probability of myocardial infarction" are made.

[0574] Output: Health risk probability information (JSON format).

[0575] Step 6:

[0576] The device displays the estimated results.

[0577] Input: Probability information of health risks.

[0578] Processing: Visually displaying received information in the form of lists and graphs, providing information in a format that is easy for users to understand.

[0579] Output: Visually displayed estimation results.

[0580] Step 7:

[0581] The device recognizes the emotional state.

[0582] Input: The user's facial expressions, tone of voice, and input content when operating the device.

[0583] Processing: Emotion recognition means are used to perform natural language processing and facial expression analysis to assess the user's emotional state.

[0584] Output: Perceived emotional state (e.g., stressed, relieved).

[0585] Step 8:

[0586] The server customizes how the information is presented.

[0587] Input: Perceived emotional state.

[0588] Treatment: Based on the emotional state, determine how to present information in a way that is most acceptable to the user. For example, use gentle language and detailed instructions for a stressed user.

[0589] Output: A customized way to provide information.

[0590] Step 9:

[0591] The device will provide customized advice.

[0592] Input: Customized information delivery method.

[0593] Processing: Explains health risks and preventative measures to the user in detail and displays auto-generated documents such as medical certificates and prescriptions in an editable format.

[0594] Output: Customized advice and editable documentation.

[0595] Step 10:

[0596] The server stores the data and trains the machine learning model.

[0597] Input: Final diagnosis results and countermeasure results.

[0598] Processing: The data is stored in a database and used to train a machine learning model, which improves the accuracy of the next analysis.

[0599] Output: A system with improved accuracy.

[0600] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0601] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0602] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0603] [Second embodiment]

[0604] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0605] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0606] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0607] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0608] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0609] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0610] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0611] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0612] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0613] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0614] In the smart glasses 214, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0615] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0616] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and assists them in estimating disease probability and creating documents based on that information. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means.

[0617] 1. Entering patient information and initial processing

[0618] User

[0619] The user (healthcare worker) logs in to a dedicated terminal (PC, tablet, etc.) and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, current test results, etc. Test results can be captured using a scanner or entered directly into the terminal.

[0620] Terminal

[0621] The device converts the input information into a standard format and sends it to the server, encrypting the data to prevent it from leaking, and notifying the user that the data has been sent.

[0622] 2. Data Acquisition and Analysis

[0623] server

[0624] The server stores the patient information sent from the device in a database. It also performs a preliminary data consistency check. It automatically generates keywords for related symptoms based on the patient information. This keyword generation is performed using natural language processing (NLP) to extract key keywords from symptoms and test results.

[0625] Based on the generated keywords, the server searches major literature databases and case report sites on the Internet, using medical database APIs and search engine queries. Related cases and literature are retrieved from the search results, and the results are filtered to extract the most relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[0626] 3. Probability estimation and information provision

[0627] server

[0628] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, "The probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%." These estimated results are converted into JSON format and sent to the terminal.

[0629] Terminal

[0630] The device visually displays the received information so that the user can check it. Possible display formats include lists and graphs. The device also automatically generates documents such as medical certificates, prescriptions, and referral letters based on the inference results. The generated documents are displayed in an editable format so that the user can check and modify them.

[0631] 4. Document Creation and Editing

[0632] server

[0633] The server prepares templates for medical certificates and prescriptions based on the inference results, and automatically generates draft documents by inserting the patient's individual information. This document generation is based on predefined templates.

[0634] Terminal

[0635] The terminal displays the generated draft document to the user and provides an interface that allows for corrections and additions as necessary. The user then checks and corrects the document and finalizes it based on the results.

[0636] 5. Data accumulation and accuracy improvement

[0637] server

[0638] The server stores the final diagnosis and treatment progress in a database. The stored data is then used to train the machine learning model. The accuracy of the system is improved by periodically retraining the machine learning model with new data sets.

[0639] As a specific example, if a patient complains of "headache, fever, cough," the user enters this information into the terminal and the information is sent to the server. The server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, this information is sent to the terminal for the user to confirm. Furthermore, the necessary medical certificates and prescriptions are automatically generated based on this information, and the user makes the final confirmation.

[0640] In this way, the system can provide information quickly and accurately, reducing the burden on medical professionals.

[0641] The processing flow will be explained below.

[0642] Step 1: Enter patient information

[0643] User

[0644] The user (medical worker) logs in to a dedicated terminal.

[0645] The user enters the patient's basic information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results.

[0646] Step 2: Sending data

[0647] Terminal

[0648] The terminal converts the entered patient information into a standard format.

[0649] The terminal transmits information to the server using encrypted communication.

[0650] The terminal notifies the user that the transmission is complete.

[0651] Step 3: Save data and check its integrity

[0652] server

[0653] The server stores the received patient information in a database.

[0654] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[0655] Step 4: Generate keywords

[0656] server

[0657] The server analyzes the stored patient information and generates keywords for relevant symptoms.

[0658] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[0659] Step 5: Internet Search

[0660] server

[0661] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[0662] Use medical database APIs or search engine queries to retrieve relevant case information.

[0663] Step 6: Filtering case information

[0664] server

[0665] The server performs filtering processing to extract highly relevant information from the acquired case information.

[0666] Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[0667] Step 7: Estimate the probability of disease

[0668] server

[0669] The server performs statistical analysis based on the filtered case information.

[0670] The probability of each disease is estimated, for example, "40% probability of influenza," "30% probability of cold," and "20% probability of pneumonia."

[0671] Step 8: Submit your information

[0672] server

[0673] The server converts the estimation results into JSON format and sends them to the terminal.

[0674] Step 9: Viewing Information

[0675] Terminal

[0676] The device visually displays the received information, for example in the form of a list or graph.

[0677] The user reviews and evaluates the displayed information.

[0678] Step 10: Prepare for document generation

[0679] server

[0680] The server prepares templates for medical certificates, prescriptions, referral letters, etc. based on the estimation results.

[0681] Insert the patient's individual information into the template and generate a draft document.

[0682] Step 11: Editing the Document

[0683] Terminal

[0684] The terminal displays the generated draft document to the user and provides it in an editable format.

[0685] The user checks the draft document and makes any necessary corrections.

[0686] Step 12: Save your data

[0687] server

[0688] The server stores the final diagnosis results and treatment progress in a database.

[0689] Step 13: Improve the accuracy of the system

[0690] server

[0691] The server uses the stored data to train the machine learning model.

[0692] The model is periodically retrained with new datasets to improve the accuracy of the system.

[0693] Example 1

[0694] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0695] In conventional medical information processing systems, medical professionals had to manually input patient information and then create diagnoses and documents based on that information, which was a cumbersome process. It also required a great deal of time and effort to efficiently search for related case information based on individual patient information and obtain appropriate diagnostic results. Furthermore, the system lacked the functionality to automatically generate and edit documents based on diagnostic results, increasing the burden on medical professionals. In addition, improving the accuracy of machine learning models also required manual effort, making the automation and accuracy improvement of the entire system a challenge.

[0696] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0697] In this invention, the server includes an information input means for inputting patient information, a keyword generation means for generating keywords based on the input patient information, a data acquisition means for acquiring case information from an external database using the generated keywords, a probability estimation means for estimating the probability of disease based on the acquired case information, an information display means for displaying the estimated probability information, a document generation means for automatically generating a medical certificate or prescription based on the estimation result, a document editing means for displaying the document generated by the document generation means in a format that can be edited by the user, and a learning means for improving the accuracy of the system using stored data. This reduces the burden on medical professionals and enables efficient and accurate diagnosis and document creation.

[0698] "Patient information" is a general term for a patient's basic personal information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[0699] "Information input means" refers to devices and software that allow a user to input patient information, and includes terminals and special input software.

[0700] "Keyword generation means" refers to a means for automatically extracting key keywords from related symptoms and test results based on input patient information.

[0701] "Data acquisition means" refers to a means for searching and acquiring case information from external databases or the Internet using the generated keywords.

[0702] "Probability estimation means" refers to a means of estimating the probability of disease occurrence based on acquired case information using statistical analysis and machine learning techniques.

[0703] "Information display means" refers to a device or software for visually displaying estimated probability information to a user, including display in list or graph format.

[0704] "Document generation means" refers to software or algorithms for automatically generating medical certificates and prescriptions based on the inference results.

[0705] The "document editing means" refers to a means for providing an interface that allows a user to modify or add to a generated document.

[0706] "Learning means" refers to the means of training a machine learning model using stored patient information and diagnostic results to improve the accuracy of the system.

[0707] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and, based on that information, assists in disease probability estimation and document creation. The system includes an information input means for inputting patient information, a keyword generation means for generating keywords, a data acquisition means for searching case information, a probability estimation means for analyzing the acquired information, an information display means for displaying the analysis results, a document generation means for automatically generating documents, a document editing means for editing the generated documents, and a learning means for improving the accuracy of the system.

[0708] Hardware and Software Configuration

[0709] Information input method

[0710] User:

[0711] Users (healthcare workers) log in to a dedicated terminal (PC, tablet, etc.), which is equipped with an easy-to-use GUI (graphical user interface), and enter basic patient information (name, age, gender, etc.) and specific symptoms (e.g., headache, fever, cough).

[0712] Device:

[0713] The device has a software module that converts the information into a standard format, such as JSON or XML, and then the converted data is encrypted and securely sent to the server.

[0714] Probability estimation means and information display means

[0715] server:

[0716] The server stores the received patient information in a database and performs an initial consistency check. It then uses natural language processing (NLP) technology to extract relevant keywords from the data. The generated keywords are then used to search for related cases and literature using medical database APIs and search engines. After filtering, the most reliable information is used to estimate the probability of disease using statistical analysis and machine learning algorithms.

[0717] Device:

[0718] The terminal visually displays the estimation results, which can be displayed in list or graph format for the user to review.

[0719] Document creation and editing tools

[0720] server:

[0721] The server prepares templates for medical certificates and prescriptions based on the estimation results, and automatically generates draft documents by inserting the patient's individual information. These templates are predefined.

[0722] Device:

[0723] The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections or additions. The user can then check, correct, and finalize the document.

[0724] Learning tools

[0725] server:

[0726] The server uses the stored data to train the machine learning model, which is periodically retrained with new data sets to improve the system's accuracy. This retraining process uses advanced machine learning techniques such as deep learning.

[0727] Specific examples

[0728] For example, if a patient complains of "headache, fever, cough," the following process occurs:

[0729] User:

[0730] Symptoms are entered into the terminal, which converts them into a standard format and sends them to the server via encrypted communication.

[0731] server:

[0732] It receives information, generates keywords such as "headache," "fever," and "cough," and searches and filters related cases and literature. For example, it estimates the probability of influenza at 40%, cold at 30%, and pneumonia at 20%.

[0733] Device:

[0734] The system displays the estimated results and automatically generates the necessary medical certificates and prescriptions, which the user can then review and finalize.

[0735] Prompt Sentence Examples

[0736] "The patient's symptoms are headache, fever, and cough. Past medical history includes allergic rhinitis, and current test results show an elevated white blood cell count. Based on this information, generate search keywords and related case information, and calculate the estimated probability of the most relevant disease. Additionally, create a draft medical certificate based on the results."

[0737] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0738] Step 1:

[0739] The user enters information

[0740] Input: Patient basic information (name, age, gender, etc.), specific symptoms (e.g. headache, fever, cough), past medical history, and current test results.

[0741] How it works: The user (healthcare worker) logs in to a dedicated device (PC, tablet, etc.) and enters patient information. The user enters information using an easy-to-use GUI.

[0742] Output: The entered patient information is temporarily saved in the terminal's memory.

[0743] Step 2:

[0744] The device prepares the data

[0745] Input: Patient information entered by the user.

[0746] How it works: The device converts the input information into a standard format (e.g., JSON or XML) to facilitate further processing. The converted data is then encrypted and security protocols are applied to prevent external access.

[0747] Output: The encrypted data converted into a standard format.

[0748] Step 3:

[0749] The device sends the data to the server

[0750] Input: Encrypted data converted to a standard format.

[0751] Operation: The device sends encrypted data to the server. The communication protocol is SSL / TLS, and encryption is used to prevent unauthorized access or tampering of data. After the transmission is complete, the device notifies the user that the transmission was successful.

[0752] Output: Data sent to the server and notification to the user that the data has been sent.

[0753] Step 4:

[0754] The server receives and stores the data

[0755] Input: Encrypted data sent from the terminal.

[0756] Operation: The server decrypts the received data and stores it in the database. A relational database is used as the database system. Before storing it in the database, the data is checked for consistency, and an error log is recorded if there is an inconsistency.

[0757] Output: Patient information stored in database and error log (if required).

[0758] Step 5:

[0759] The server generates the keywords

[0760] Input: Patient information stored in the database.

[0761] How it works: The server uses natural language processing (NLP) techniques to extract key keywords from the patient information entered. For example, if the patient information contains "headache," "fever," and "cough," those keywords are generated.

[0762] Output: The generated keywords.

[0763] Step 6:

[0764] The server searches for case information

[0765] Input: Generated keywords.

[0766] How it works: The server uses the generated keywords to search for case reports and literature on the Internet using medical database APIs and search engines. It queries specific databases (e.g., medical research databases or academic paper databases) to obtain relevant information. The obtained data is then filtered to select information that is highly reliable and closely matches the keywords.

[0767] Output: Filtered case information.

[0768] Step 7:

[0769] The server estimates the probability of the disease

[0770] Input: Filtered case information.

[0771] How it works: The server uses statistical analysis and machine learning algorithms to estimate disease probabilities based on filtered case information, such as "the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%."

[0772] Output: Estimated probability information.

[0773] Step 8:

[0774] The server sends the estimation results to the device.

[0775] Input: Estimated probability information.

[0776] How it works: The server converts the probability information into JSON format and sends it to the device. The transmission process is secure using the SSL / TLS protocol.

[0777] Output: Probability information sent to the terminal.

[0778] Step 9:

[0779] The device displays the estimated results.

[0780] Input: Probability information sent to the terminal.

[0781] Operation: The device visually displays the received probability information in a list or graph format designed to be easily understood by the user.

[0782] Output: Probability information displayed to the user.

[0783] Step 10:

[0784] The server automatically generates the document

[0785] Input: Estimated probability information and basic patient information.

[0786] How it works: The server prepares templates for medical certificates and prescriptions based on the inference results, embeds the patient's individual information, and automatically generates draft documents using a template engine.

[0787] Output: Auto-generated draft document.

[0788] Step 11:

[0789] The device edits and displays the document.

[0790] Input: Auto-generated draft document.

[0791] Operation: The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections and additions. Once editing is complete, the user performs a final check and finalizes the document.

[0792] Output: Final revised and verified document.

[0793] Step 12:

[0794] The server stores the data and updates the learning model.

[0795] Input: Finalized documentation and diagnostic results.

[0796] How it works: The server stores the final diagnosis and treatment progress in a database. The stored data is used to periodically retrain the machine learning model to improve the accuracy of the system. The machine learning algorithm includes adaptive learning to accommodate increasing data.

[0797] Output: Updated machine learning model and saved diagnostic data.

[0798] (Application example 1)

[0799] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0800] In modern factories, maintenance and inspection of each machine is still often done manually. This relies on the knowledge and experience of the maintenance personnel, which can delay early detection of abnormalities and optimal maintenance responses. Furthermore, updating maintenance histories and manuals is time-consuming, which tends to lower the efficiency of the entire factory. Furthermore, identifying the cause of an abnormality and selecting a countermeasure can take time, so an efficient system to solve these problems is needed.

[0801] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0802] In this invention, the server includes input means for inputting patient information, means for generating keywords based on the input patient information, search means for retrieving case information from an external database using the generated keywords, analysis means for estimating the probability of disease based on the retrieved case information, means for displaying the estimated probability information, input means for inputting each device information, means for generating keywords based on the input device information, search means for retrieving maintenance response actions from an external database using the generated keywords, and means for applying the retrieved maintenance response actions to machine maintenance and repair. This makes it possible to efficiently and quickly identify machine abnormalities and propose optimal maintenance responses.

[0803] "Patient information" refers to basic patient data (such as name, age, sex, past medical history, specific symptoms, and test results) entered by medical professionals.

[0804] "Input means" refers to equipment or devices (PCs, tablets, smartphones, etc.) used to collect patient information and information on machinery and equipment and input it as data.

[0805] "Keyword generation means" refers to the technology or algorithm used to extract key keywords from input information.

[0806] A "search means" is a system or software that uses the generated keywords to retrieve related information from an external database.

[0807] "Analysis means" refers to a device or program that has the function of estimating and displaying specific results using statistical analysis or machine learning based on acquired information.

[0808] "Display means" refers to a display or monitor for visually displaying analysis results, estimated probability information, maintenance response actions, etc.

[0809] "Each device information" refers to data on the status and abnormalities of various machinery and equipment within the factory (such as years of operation, error logs, and current operating status).

[0810] "Maintenance response actions" refer to specific procedures and tasks for resolving abnormalities or breakdowns in machinery and equipment.

[0811] "Document generation means" refers to software or a system for automatically generating documents such as medical certificates, prescriptions, and maintenance manuals based on acquired information.

[0812] "Learning" is the process of using stored data to train machine learning models to improve the system's analytical accuracy and predictive capabilities.

[0813] This invention relates to a "smart maintenance system" for improving the efficiency of maintenance and inspection work for factory robots and machinery equipment. In particular, it is a system for automating a series of processes including input of equipment information, keyword generation, related information search, analysis, display, document generation, and learning.

[0814] 1. Entering device information and initial processing

[0815] User

[0816] Users (maintenance personnel) use dedicated devices (smartphones, tablets, etc.) to input information about each piece of machinery and equipment. For example, they input data such as the machine's name, years of operation, current status, and error logs. This information plays an important role in subsequent analysis and maintenance measures.

[0817] Terminal

[0818] The terminal converts the input data into a standard format and transmits it to the server using secure encrypted communications. After transmitting the data, the terminal notifies the user that the transmission has been completed.

[0819] 2. Data Acquisition and Analysis

[0820] server

[0821] The server stores the machine equipment data sent from the terminal in a database. At the same time, it checks the data for consistency. Next, it uses natural language processing (NLP) to generate keywords from the input information. For example, it extracts key keywords from error logs, such as "loud operating noise" or "vibration felt."

[0822] Based on the generated keywords, the server searches an external database for relevant maintenance actions, filtering the results based on criteria such as reliable data sources and keyword match.

[0823] 3. Estimation and provision of information on maintenance response actions

[0824] server

[0825] The server analyzes the filtered information and predicts the most appropriate maintenance action. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." The predicted maintenance action is sent to the terminal in JSON format.

[0826] Terminal

[0827] The terminal visually displays the received maintenance response actions. Possible display formats include lists, graphs, and text. In addition, documents such as maintenance manuals and maintenance histories are automatically generated based on the estimation results. The generated documents are displayed in a format that can be checked and edited by the user.

[0828] 4. Document Creation and Editing

[0829] server

[0830] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This document generation is performed based on predefined templates.

[0831] Terminal

[0832] The terminal displays the generated draft document to the user and provides an editing interface that allows the user to make corrections or additions as necessary. After the user checks and corrects the document, it is finalized.

[0833] 5. Data accumulation and accuracy improvement

[0834] server

[0835] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved.

[0836] Specific examples

[0837] The maintenance technician enters the status information of the "Robot Arm A1" (2 years of operation, operating noise, vibration detection) into the terminal. The input data is converted to JSON format and sent to the maintenance server.

[0838] The server analyzes the error log information sent and extracts key keywords ("operating sound" and "vibration") using TF-IDF. Maintenance response actions are searched and filtered based on the obtained keywords.

[0839] Based on the extracted keywords, the most appropriate maintenance actions ("Check the cooling system and clean the filters" and "Check for loose parts and lubricate the bearings") are displayed.

[0840] Prompt Sentence Examples

[0841] "You input the machine's condition, for example, 'it's loud when it's running' or 'I feel vibrations.' Then, based on this information, we suggest the best maintenance action."

[0842] As described above, the present invention provides a system that improves the efficiency of maintenance management within a factory, reduces the burden on maintenance personnel, and enables early detection of machine abnormalities.

[0843] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0844] Step 1:

[0845] Users input information about each piece of machinery and equipment using a dedicated device (smartphone, tablet, etc.). The input includes the machine's name, years of operation, current status, error log, etc. This allows data input (machine information) from the user to be obtained. This data is the basic data required for subsequent analysis.

[0846] Step 2:

[0847] The terminal converts the machine information entered by the user into a standard format, for example, JSON format, and sends it to the server using encrypted communication. This allows data to be sent from the terminal to the server.

[0848] Step 3:

[0849] The server receives the machine information sent from the terminal and stores it in a database. Before storing it, it checks the data for consistency and ensures that there are no abnormalities. This ensures the reliability of the data.

[0850] Step 4:

[0851] The server performs natural language processing (NLP) based on the received machine information to extract key keywords. For example, keywords such as "loud operating noise" and "vibration felt" are generated from error logs using TF-IDF. This provides basic keywords for analysis.

[0852] Step 5:

[0853] The server queries an external database (such as a maintenance manual database) based on the generated keywords. It extracts appropriate maintenance response actions and collects information on them. Search results are filtered based on reliable data sources, keyword matching, etc.

[0854] Step 6:

[0855] The server performs statistical analysis on the acquired maintenance response actions and estimates specific maintenance measures. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." This allows it to estimate the optimal maintenance response action.

[0856] Step 7:

[0857] The server converts the estimated maintenance action into JSON format and sends it to the terminal, which then transmits the analysis results to the terminal.

[0858] Step 8:

[0859] The terminal visually displays the received maintenance action and notifies the user. The display format can be a list, graph, text, etc. This allows the user to confirm the maintenance action.

[0860] Step 9:

[0861] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This generates documents for maintenance actions.

[0862] Step 10:

[0863] The terminal displays the generated draft document to the user and provides an editing interface that allows corrections and additions as necessary. After the user checks and corrects the document, they finalize it. This finalizes the maintenance document.

[0864] Step 11:

[0865] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved, thereby improving overall system performance.

[0866] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0867] The present invention relates to an information provision system that efficiently inputs patient information and, based on that information, estimates disease probability, creates documents, and recognizes user emotions. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means. Furthermore, it incorporates an emotion engine and has the function of customizing the method of providing diagnosis results and advice based on the user's emotional state.

[0868] 1. Entering patient information and initial processing

[0869] User

[0870] The user (healthcare worker) logs in to a dedicated terminal and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly into the terminal.

[0871] Terminal

[0872] The terminal converts the input information into a standard format and transmits it to the server using encrypted communications. Once transmission is complete, the terminal notifies the user.

[0873] 2. Data Acquisition and Analysis

[0874] server

[0875] The server stores the patient information sent from the terminal in a database, checks the data consistency, and sends a supplemental request to the terminal if any information is missing.

[0876] server

[0877] The server analyzes stored patient information and generates keywords for related symptoms, using natural language processing (NLP) to extract key keywords from symptoms and test results.

[0878] server

[0879] The server searches online literature databases and case report sites based on the generated keywords, and retrieves relevant case information using medical database APIs and search engine queries.

[0880] server

[0881] The server performs filtering to extract highly relevant information from the acquired case information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[0882] server

[0883] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, it may estimate "40% probability of influenza," "30% probability of cold," or "20% probability of pneumonia." These estimation results are converted into JSON format and sent to the device.

[0884] 3. Information provision and document generation

[0885] Terminal

[0886] The device visually displays the received information so that the user can check it. It can use display formats such as lists and graphs. Furthermore, it automatically generates medical certificates, prescriptions, referral letters, etc. based on the inference results.

[0887] server

[0888] Based on the estimation results, the server prepares templates for medical certificates and prescriptions, inserts the patient's individual information, and generates draft documents.

[0889] Terminal

[0890] The terminal provides the generated draft document in an editable format, allowing the user to check and modify the draft document and make any necessary additions.

[0891] 4. Emotion Recognition and Response

[0892] Terminal

[0893] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[0894] server

[0895] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[0896] server

[0897] Based on the assessment, the server adjusts the presentation of the diagnostic results to provide information in a format that is most acceptable to the user, for example, using calmer, more reassuring language for a user in a stressed state.

[0898] Terminal

[0899] The device uses an emotion engine to provide users with customized medical advice, for example, providing detailed instructions on specific steps to take or the need for additional testing if the user is worried.

[0900] 5. Data accumulation and learning

[0901] server

[0902] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[0903] As a concrete example, if a patient complains of "headache, fever, and cough," the user enters this information into the terminal, and the server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, the terminal will display this information and further analyze the user's emotional state to provide information in the most optimal form. Medical certificates and prescriptions are also automatically generated, with the user making the final confirmation.

[0904] This system can provide information quickly and accurately, reducing the burden on medical professionals. Furthermore, emotion recognition can be used to present information in a way that is easy for users to accept, improving the quality of medical care.

[0905] The processing flow will be explained below.

[0906] Step 1: Enter patient information

[0907] User

[0908] The user (medical worker) logs in to a dedicated terminal.

[0909] The user enters the patient's basic information (name, age, gender, etc.), symptoms (headache, fever, cough, etc.), past medical history, and current test results.

[0910] Step 2: Sending data

[0911] Terminal

[0912] The terminal converts the entered patient information into a standard format.

[0913] The terminal transmits the patient information to the server using encrypted communication.

[0914] The terminal notifies the user that the transmission is complete.

[0915] Step 3: Save data and check its integrity

[0916] server

[0917] The server stores the received patient information in a database.

[0918] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[0919] Step 4: Generate keywords

[0920] server

[0921] The server analyzes stored patient information and automatically generates keywords for relevant symptoms.

[0922] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[0923] Step 5: Internet Search

[0924] server

[0925] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[0926] Use medical database APIs or search engine queries to retrieve relevant case information.

[0927] Step 6: Filtering case information

[0928] server

[0929] The server performs filtering to extract highly relevant information from the acquired case information.

[0930] Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[0931] Step 7: Estimate the probability of disease

[0932] server

[0933] The server performs statistical analysis based on the filtered case information and estimates the probability of each disease.

[0934] For example, estimates might say, "The probability of influenza is 40%," "The probability of a cold is 30%," and "The probability of pneumonia is 20%."

[0935] Step 8: Submit your information

[0936] server

[0937] The server converts the estimated probability information into JSON format and sends it to the terminal.

[0938] Step 9: Viewing Information

[0939] Terminal

[0940] The device visually displays the information it receives, for example in the form of a list or graph.

[0941] The user reviews and evaluates the displayed information.

[0942] Step 10: Emotion Recognition

[0943] Terminal

[0944] The device uses an emotion engine to analyze the user's facial expressions, tone of voice, and input content while operating the device, and recognizes the user's emotional state.

[0945] Step 11: Assess and regulate your emotional state

[0946] server

[0947] The server evaluates the user's emotional state (stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[0948] The server adjusts the way the information is presented based on the evaluation results, and provides the information in a format that is easy for the user to accept.

[0949] Step 12: Customize your information offering

[0950] Terminal

[0951] The terminal uses the emotion engine to present customized medical advice to the user.

[0952] For example, a user in a good emotional state may be provided with concise and efficient information, while a user experiencing stress may be provided with detailed and reassuring information.

[0953] Step 13: Document Generation

[0954] server

[0955] The server prepares templates for a diagnosis, prescription, and referral letter based on the estimation results and the user's emotional state.

[0956] Insert the patient's individual information into the template and automatically generate a draft document.

[0957] Step 14: Edit and review the document

[0958] Terminal

[0959] The terminal displays the generated draft document to the user and provides it in an editable format.

[0960] The user checks and corrects the draft document and makes any necessary additions.

[0961] Step 15: Save your data

[0962] server

[0963] The server stores the final diagnosis results and treatment progress in a database.

[0964] Step 16: Training the system and improving accuracy

[0965] server

[0966] The server uses the stored data to train the machine learning model.

[0967] The model is periodically retrained with new datasets to improve the accuracy of the system.

[0968] Example 2

[0969] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0970] In modern medical settings, entering and analyzing patient information, estimating diagnostic probabilities, and determining optimal treatment plans requires a great deal of time and effort. Furthermore, there is often a lack of information provided that takes into account the patient's emotions and mental state, which increases the burden on medical professionals and increases patient anxiety. To efficiently resolve these issues and improve the quality of medical care, a comprehensive and flexible information provision system is required.

[0971] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0972] In this invention, the server includes means for inputting patient information, means for converting the input patient information into a standard format and transmitting it, means for saving the transmitted patient information, means for generating keywords based on the saved patient information, means for retrieving case information from an external database using the generated keywords, means for filtering the retrieved case information based on relevance, means for estimating disease probability based on the filtered case information, means for displaying the estimated probability information, and means for analyzing user emotions and adjusting the information presentation method, thereby enabling efficient input and analysis of patient information, accurate disease probability estimation, and information provision that takes patient emotions into consideration.

[0973] "Patient information" is a general term for basic patient information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[0974] "Means for inputting" refers to a device or interface that allows a user to input patient information.

[0975] "Means for converting into a standard format and transmitting" refers to the function for converting input patient information into a specified format and transmitting it securely to a server using an encrypted communication protocol.

[0976] "Means for storing" refers to a function for storing transmitted patient information in a database.

[0977] "Means for generating keywords" refers to technology that analyzes stored patient information and extracts keywords based on key symptoms and test results.

[0978] "Means for obtaining case information" refers to a function for obtaining related case information from external databases or case report sites using the generated keywords.

[0979] "Filtering means" refers to a function for sorting acquired case information based on relevance and extracting only highly reliable information.

[0980] "Means for estimating the probability of disease" refers to technology for performing statistical analysis based on filtered case information to calculate the probability of occurrence of a specific disease.

[0981] "Means for displaying probability information" refers to a display interface that allows a user to visually confirm the estimated disease occurrence probability.

[0982] "Means for analyzing emotions and adjusting the way information is presented" refers to a function that analyzes the user's emotional state, such as facial expressions and tone of voice, and customizes the way information is presented based on the results.

[0983] "Means for automatically generating documents" refers to a function for automatically generating medical documents such as medical certificates, prescriptions, and referral letters.

[0984] "Means for displaying in an editable format" refers to an interface for displaying the generated medical document in a format that can be checked and corrected by the user.

[0985] "Means for training machine learning models to improve the accuracy of the system" refers to technology that uses stored data on past diagnostic results and treatment progress to train machine learning models and improve the diagnostic accuracy of the system.

[0986] The present invention relates to an information providing system that efficiently inputs patient information and, based on the input information, estimates the probability of disease, creates documents, and recognizes user emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[0987] System configuration and basic operations

[0988] User

[0989] The user (healthcare worker) logs in to a dedicated terminal. The terminal provides an interface for entering basic patient information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly by hand. Once the user has finished entering the patient information, the terminal converts this information into a standard format and sends it to the server using an encrypted communication protocol (e.g., HTTPS).

[0990] Terminal

[0991] The terminal converts the received patient information into a standard JSON format and sends the information to the server using the HTTPS protocol. Once the transmission is complete, a notification of the completion is displayed to the user.

[0992] server

[0993] The server stores the patient information sent from the device in a database. It checks the consistency of the stored information and sends a supplemental request to the device if any information is missing. After consistency is confirmed, the server uses natural language processing (NLP) technology (e.g., spaCy or NLTK) to generate keywords from the patient information based on key symptoms and test results. Using the generated keywords, the server retrieves relevant case information from external literature databases and case report sites via medical database APIs (e.g., PubMed, ClinicalTrials.gov) or search engines.

[0994] The server filters the acquired case information and selects only the most relevant information. Based on the filtered information, the server performs statistical analysis to estimate the probability of a specific disease occurring. The estimated results (e.g., "40% probability of influenza," "30% probability of cold," "20% probability of pneumonia") are converted to JSON format and sent back to the terminal.

[0995] Linking devices and servers

[0996] The terminal visually displays the received probability estimation results, using formats such as lists and graphs. Furthermore, based on the information sent from the server, the terminal automatically generates medical documents such as medical certificates, prescriptions, and referral letters. The generated documents are provided in an editable format that the user can review and modify.

[0997] Emotion recognition function

[0998] When the user operates the device, the device uses an emotion engine to recognize the user's emotional state, such as facial expressions and tone of voice. The server evaluates the user's emotional state based on the analysis results of the emotion engine and adjusts the way the diagnosis results are presented. For example, if the user is anxious, the system will display information in calm language and provide detailed instructions on how to deal with the situation.

[0999] Learning Function

[1000] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model (e.g., neural network) to improve the accuracy of the system.

[1001] Specific examples

[1002] For example, if a patient complains of headache, fever, and cough, the user enters this information into the device. The server uses natural language processing to automatically generate keywords such as "headache," "fever," and "cough," and searches the case database. The search results include an estimate of a 40% probability of influenza, a 30% probability of a cold, and a 20% probability of pneumonia. These estimates are sent to the device in JSON format, where they are displayed visually in graphs. The emotion engine analyzes the user's facial expressions and provides customized advice, such as displaying the diagnosis results in calm language. Furthermore, a medical certificate and prescription are automatically generated, and the user makes the final confirmation.

[1003] Example prompts for generative AI models

[1004] "A patient complains of headache, fever, and cough. Their only past medical history is allergic rhinitis. Estimate the probability of influenza and visually display the results. Also, if the user is stressed, please respond by providing information in a calmer tone."

[1005] This system enables prompt and accurate information provision, reducing the burden on medical professionals. In addition, the emotion recognition function allows information to be provided in a format that is easy for users to accept, improving the quality of medical care.

[1006] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1007] Step 1:

[1008] User login and patient information entry

[1009] User

[1010] The user logs in to a dedicated terminal. They access the login screen, enter their user ID and password, and click the login button. If login is successful, they are redirected to the patient information input screen. The user enters the patient's basic information (name, age, sex), specific symptoms (headache, fever, cough, etc.), past medical history, and current test results in the form on the screen, and clicks the "Submit" button.

[1011] Input and Output

[1012] Input: Patient information entered by the user into the terminal

[1013] Output: The input information is converted into a standard format.

[1014] Step 2:

[1015] Converting and sending input information

[1016] Terminal

[1017] The terminal converts the received patient information into a standard format (e.g., JSON format) and sends it to the server using an encrypted communication protocol (e.g., HTTPS). Once the transmission is complete, the terminal displays a pop-up message stating, "Information transmission completed."

[1018] Input and Output

[1019] Input: User-entered patient information

[1020] Output: Information converted to standard format and notification of completion of transmission

[1021] Step 3:

[1022] Server information storage and supplementary requests

[1023] server

[1024] The server receives the patient information sent from the terminal and stores it in the database. After storing it, it checks the consistency of the data and sends a supplementary request to the terminal if necessary information is missing.

[1025] Input and Output

[1026] Input: Patient information submitted in a standard format

[1027] Output: Stored information and supplementary requests after consistency check

[1028] Step 4:

[1029] Generate keywords for patient information

[1030] server

[1031] The server analyzes the stored patient information and generates keywords based on relevant symptoms and test results using natural language processing (NLP) techniques (e.g., spaCy and NLTK).

[1032] Input and Output

[1033] Input: Saved patient information

[1034] Output: Generated keyword list

[1035] Step 5:

[1036] Case database search

[1037] server

[1038] The server searches external literature databases and case report sites based on the generated keywords, using medical database APIs and search engine queries.

[1039] Input and Output

[1040] Input: Generated keyword list

[1041] Output: Retrieved case information

[1042] Step 6:

[1043] Case information filtering

[1044] server

[1045] The server then performs a filtering process based on the acquired case information to extract highly relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[1046] Input and Output

[1047] Input: Retrieved case information

[1048] Output: Filtered case information list

[1049] Step 7:

[1050] Disease probability estimation and result transmission

[1051] server

[1052] The server performs statistical analysis based on the filtered case information to estimate the probability of candidate diseases. The estimated results are converted into JSON format and sent to the device.

[1053] Input and Output

[1054] Input: Filtered case information

[1055] Output: Estimated disease probability information

[1056] Step 8:

[1057] Displaying diagnostic results and generating documentation

[1058] Terminal

[1059] The device visually displays the received information for the user to review. Display formats include lists and graphs. Furthermore, medical documents such as medical certificates, prescriptions, and referral letters are automatically generated based on the inference results. The generated documents are provided in an editable format, allowing the user to review and correct them.

[1060] Input and Output

[1061] Input: Estimated disease probability information

[1062] Output: Visually displayed information and generated medical documents

[1063] Step 9:

[1064] Emotion recognition and information presentation adjustment

[1065] Terminal

[1066] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[1067] server

[1068] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine. Based on the evaluation results, it adjusts the way the diagnosis results are presented to provide information in a format that is most acceptable to the user. For example, for a user in a stressed state, it uses calmer, more reassuring language.

[1069] Terminal

[1070] The device uses an emotion engine to provide users with customized medical advice, detailing specific steps to take or the need for additional testing if the user is concerned.

[1071] Input and Output

[1072] Input: User's emotional state

[1073] Output: customized diagnostic results

[1074] Step 10:

[1075] Data accumulation and learning

[1076] server

[1077] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[1078] Input and Output

[1079] Input: Final diagnosis and treatment progress

[1080] Output: Updated machine learning model

[1081] Specific actions

[1082] In this way, through a series of processes, patient information is efficiently input and analyzed, appropriate disease probability is estimated, and optimal treatment plans are determined.In addition, by presenting information that takes the user's emotions into consideration, the burden on medical professionals is reduced and patients feel more at ease.

[1083] (Application example 2)

[1084] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1085] Conventional health management systems for factory workers have difficulty efficiently integrating multiple health data sets and quickly and accurately assessing health risks. Furthermore, they do not provide appropriate advice based on the worker's emotional state, resulting in decreased productivity and the risk of workers' health deteriorating. These issues needed to be resolved.

[1086] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1087] In this invention, the server includes an input means for inputting the health data of workers, a means for generating keywords based on the input health data, a search means for retrieving case information from an external database using the generated keywords, an analysis means for estimating the probability of a health risk based on the retrieved case information, a means for displaying the estimated probability information, an emotion recognition means for recognizing the emotional state of the user, a means for customizing the information provision method based on the recognized emotional state, a document generation means for automatically generating documents such as medical certificates and prescriptions, a means for displaying the generated documents in an editable format, and a means for estimating health risks based on the health data and suggesting preventive measures. This makes it possible to efficiently monitor and evaluate the health status of factory workers and provide appropriate advice according to their emotional state.

[1088] "Patient information" refers to basic information about an individual patient collected by a medical professional, including, specifically, data such as the patient's name, age, gender, symptoms, medical history, and current test results.

[1089] "Input means" refers to an interface or device that allows a user to input patient information into a terminal.

[1090] The "keyword generating means" is a means for extracting major symptoms and related information from input patient information and generating search keywords.

[1091] The "search means" is a means for retrieving related case information from an external database using the generated keywords.

[1092] The "analysis means" is a means for estimating the probability of a particular disease or health risk based on the acquired case information.

[1093] The "display means" is a means for visually displaying estimated probability information and diagnostic results in a format that can be confirmed by the user.

[1094] The "emotion recognition means" is a means for analyzing the user's facial expressions and tone of voice to recognize the user's emotional state.

[1095] The "information presentation method customization means" is a means for optimizing the information presentation method based on the recognized emotional state.

[1096] The "document generation means" is a means for automatically creating documents such as medical certificates, prescriptions, and referral letters.

[1097] "Means for displaying in an editable format" refers to means for providing the generated document in a format that allows the user to check and modify it.

[1098] "Healthcare data" refers to data that indicates the health status of factory workers, specifically including blood pressure, heart rate, symptoms, medical history, etc.

[1099] "Means for estimating health risks" are means for assessing the need for preventive measures and countermeasures based on health data and for calculating the probability of risk.

[1100] A "means for proposing preventive measures" is a means for proposing specific preventive measures or countermeasures based on estimated health risks.

[1101] This invention is applied to a health management system for factory workers. Its basic components include an input means for inputting health data, an analysis means for standardizing and analyzing the data, a display means for displaying and editing estimation results and documents, and an emotion recognition means for recognizing emotional states. It also includes a means for assessing health risks and suggesting individual preventive measures.

[1102] 1. Health data entry and initial processing

[1103] User

[1104] Users (factory workers or managers) use dedicated input devices (e.g., PCs or tablets) to enter health data into the system. This data includes blood pressure, heart rate, symptoms, medical history, etc. After completing the data entry, the user sends the data from their device to the server. At this time, the data is converted into a standard format and sent to the server using encrypted communication.

[1105] 2. Data Acquisition and Analysis

[1106] server

[1107] The server receives the health data sent from the device and stores it in a database. Natural language processing (NLP) is used to generate keywords from the stored data, and relevant case information is searched for in external databases. Search methods include medical database APIs and search engine queries.

[1108] server

[1109] The server filters the acquired case information and estimates health risks based on reliable information. For example, it may estimate "40% probability of overwork," "30% probability of high blood pressure," or "20% probability of myocardial infarction." The estimated results are converted into JSON format and sent to the device.

[1110] 3. Information provision and document generation

[1111] Terminal

[1112] The device visually displays the received information for the user to review, and can use lists and graphs as display formats. It can also automatically generate and display documents such as medical certificates, prescriptions, and referral letters in user-editable formats.

[1113] 4. Emotion Recognition and Response

[1114] Terminal

[1115] When the user operates the terminal, the terminal uses emotion recognition means to recognize the user's emotional state from the user's facial expression, tone of voice, input contents, etc.

[1116] server

[1117] Based on the analysis results of the emotion recognition means, the server evaluates the user's emotional state (e.g., stress, relief, fatigue, etc.) and optimizes the way it provides information. For example, if the user is fatigued, it will provide detailed, specific instructions on how to deal with the situation in a gentle manner.

[1118] Terminal

[1119] The device uses emotion recognition to provide customized advice to the user, detailing specific actions to take or the need for further testing if the user is concerned.

[1120] 5. Data accumulation and learning

[1121] server

[1122] The server stores the final diagnosis results and countermeasure results in a database and uses the stored data to train the machine learning model, thereby improving the accuracy of the system.

[1123] Examples:

[1124] For example, suppose Mr. Tanaka (not his real name) complains of fatigue, headache, and muscle pain during a health checkup. This information is entered into the terminal. Based on this, the server stores the keywords "fatigue," "headache," and "muscle pain," and retrieves case information from an external database. As a result, it estimates that there is a "40% chance of overwork," a "30% chance of high blood pressure," and a "20% chance of myocardial infarction," and Mr. Tanaka's emotional state is also analyzed. If Mr. Tanaka is feeling stressed, it gently explains how to deal with the situation.

[1125] Example prompt sentence:

[1126] "Design a system to monitor and analyze the health status of factory workers. Analyze the entered health data (e.g., symptoms, medical history, current test results) and estimate health risks. Additionally, analyze the user's emotional state and provide individually customized advice."

[1127] In this way, by using this system, it is possible to efficiently manage the health of factory workers and provide appropriate advice according to the emotional state of each individual worker.

[1128] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1129] Step 1:

[1130] The user enters health data.

[1131] Input: The user inputs health data such as blood pressure, heart rate, symptoms, and medical history into the device.

[1132] Processing: The entered data is converted into a standard format and sent to the server using encrypted communications.

[1133] Output: A confirmation of receipt of the data sent is displayed on the terminal.

[1134] Step 2:

[1135] The server receives and stores the data.

[1136] Input: Health data sent from the device in a standard format.

[1137] Processing: Stores the received data in a database, checks the data consistency, and sends a supplemental request to the terminal if there is missing information.

[1138] Output: The saved data. If additional information is required, the request is notified to the terminal.

[1139] Step 3:

[1140] The server generates the keywords.

[1141] Input: Stored health data.

[1142] Processing: Using natural language processing (NLP), key keywords are automatically extracted. For example, "fatigue," "headache," and "muscle pain" are extracted.

[1143] Output: The generated keywords.

[1144] Step 4:

[1145] The server retrieves case information from an external database.

[1146] Input: Generated keywords.

[1147] Processing: Search for case information based on keywords using medical database APIs and search engine queries.

[1148] Output: Retrieved case information.

[1149] Step 5:

[1150] The server estimates the health risk.

[1151] Input: Retrieved case information.

[1152] Processing: Filtering is performed and statistical analysis is carried out based on highly reliable information. For example, estimates such as "40% probability of overwork," "30% probability of high blood pressure," and "20% probability of myocardial infarction" are made.

[1153] Output: Health risk probability information (JSON format).

[1154] Step 6:

[1155] The device displays the estimated results.

[1156] Input: Probability information of health risks.

[1157] Processing: Visually displaying received information in the form of lists and graphs, providing information in a format that is easy for users to understand.

[1158] Output: Visually displayed estimation results.

[1159] Step 7:

[1160] The device recognizes the emotional state.

[1161] Input: The user's facial expressions, tone of voice, and input content when operating the device.

[1162] Processing: Emotion recognition means are used to perform natural language processing and facial expression analysis to assess the user's emotional state.

[1163] Output: Perceived emotional state (e.g., stressed, relieved).

[1164] Step 8:

[1165] The server customizes how the information is presented.

[1166] Input: Perceived emotional state.

[1167] Treatment: Based on the emotional state, determine how to present information in a way that is most acceptable to the user. For example, use gentle language and detailed instructions for a stressed user.

[1168] Output: A customized way to provide information.

[1169] Step 9:

[1170] The device will provide customized advice.

[1171] Input: Customized information delivery method.

[1172] Processing: Explains health risks and preventative measures to the user in detail and displays auto-generated documents such as medical certificates and prescriptions in an editable format.

[1173] Output: Customized advice and editable documentation.

[1174] Step 10:

[1175] The server stores the data and trains the machine learning model.

[1176] Input: Final diagnosis results and countermeasure results.

[1177] Processing: The data is stored in a database and used to train a machine learning model, which improves the accuracy of the next analysis.

[1178] Output: A system with improved accuracy.

[1179] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1180] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1181] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1182] [Third embodiment]

[1183] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1184] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1185] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1186] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1187] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1188] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1189] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1190] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1191] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1192] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1193] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1194] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1195] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and assists them in estimating disease probability and creating documents based on that information. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means.

[1196] 1. Entering patient information and initial processing

[1197] User

[1198] The user (healthcare worker) logs in to a dedicated terminal (PC, tablet, etc.) and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, current test results, etc. Test results can be captured using a scanner or entered directly into the terminal.

[1199] Terminal

[1200] The device converts the input information into a standard format and sends it to the server, encrypting the data to prevent it from leaking, and notifying the user that the data has been sent.

[1201] 2. Data Acquisition and Analysis

[1202] server

[1203] The server stores the patient information sent from the device in a database. It also performs a preliminary data consistency check. It automatically generates keywords for related symptoms based on the patient information. This keyword generation is performed using natural language processing (NLP) to extract key keywords from symptoms and test results.

[1204] Based on the generated keywords, the server searches major literature databases and case report sites on the Internet, using medical database APIs and search engine queries. Related cases and literature are retrieved from the search results, and the results are filtered to extract the most relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[1205] 3. Probability estimation and information provision

[1206] server

[1207] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, "The probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%." These estimated results are converted into JSON format and sent to the terminal.

[1208] Terminal

[1209] The device visually displays the received information so that the user can check it. Possible display formats include lists and graphs. The device also automatically generates documents such as medical certificates, prescriptions, and referral letters based on the inference results. The generated documents are displayed in an editable format so that the user can check and modify them.

[1210] 4. Document Creation and Editing

[1211] server

[1212] The server prepares templates for medical certificates and prescriptions based on the inference results, and automatically generates draft documents by inserting the patient's individual information. This document generation is based on predefined templates.

[1213] Terminal

[1214] The terminal displays the generated draft document to the user and provides an interface that allows for corrections and additions as necessary. The user then checks and corrects the document and finalizes it based on the results.

[1215] 5. Data accumulation and accuracy improvement

[1216] server

[1217] The server stores the final diagnosis and treatment progress in a database. The stored data is then used to train the machine learning model. The accuracy of the system is improved by periodically retraining the machine learning model with new data sets.

[1218] As a specific example, if a patient complains of "headache, fever, cough," the user enters this information into the terminal and the information is sent to the server. The server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, this information is sent to the terminal for the user to confirm. Furthermore, the necessary medical certificates and prescriptions are automatically generated based on this information, and the user makes the final confirmation.

[1219] In this way, the system can provide information quickly and accurately, reducing the burden on medical professionals.

[1220] The processing flow will be explained below.

[1221] Step 1: Enter patient information

[1222] User

[1223] The user (medical worker) logs in to a dedicated terminal.

[1224] The user enters the patient's basic information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results.

[1225] Step 2: Sending data

[1226] Terminal

[1227] The terminal converts the entered patient information into a standard format.

[1228] The terminal transmits information to the server using encrypted communication.

[1229] The terminal notifies the user that the transmission is complete.

[1230] Step 3: Save data and check its integrity

[1231] server

[1232] The server stores the received patient information in a database.

[1233] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[1234] Step 4: Generate keywords

[1235] server

[1236] The server analyzes the stored patient information and generates keywords for relevant symptoms.

[1237] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[1238] Step 5: Internet Search

[1239] server

[1240] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[1241] Use medical database APIs or search engine queries to retrieve relevant case information.

[1242] Step 6: Filtering case information

[1243] server

[1244] The server performs filtering processing to extract highly relevant information from the acquired case information.

[1245] Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[1246] Step 7: Estimate the probability of disease

[1247] server

[1248] The server performs statistical analysis based on the filtered case information.

[1249] The probability of each disease is estimated, for example, "40% probability of influenza," "30% probability of cold," and "20% probability of pneumonia."

[1250] Step 8: Submit your information

[1251] server

[1252] The server converts the estimation results into JSON format and sends them to the terminal.

[1253] Step 9: Viewing Information

[1254] Terminal

[1255] The device visually displays the received information, for example in the form of a list or graph.

[1256] The user reviews and evaluates the displayed information.

[1257] Step 10: Prepare for document generation

[1258] server

[1259] The server prepares templates for medical certificates, prescriptions, referral letters, etc. based on the estimation results.

[1260] Insert the patient's individual information into the template and generate a draft document.

[1261] Step 11: Editing the Document

[1262] Terminal

[1263] The terminal displays the generated draft document to the user and provides it in an editable format.

[1264] The user checks the draft document and makes any necessary corrections.

[1265] Step 12: Save your data

[1266] server

[1267] The server stores the final diagnosis results and treatment progress in a database.

[1268] Step 13: Improve the accuracy of the system

[1269] server

[1270] The server uses the stored data to train the machine learning model.

[1271] The model is periodically retrained with new datasets to improve the accuracy of the system.

[1272] Example 1

[1273] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1274] In conventional medical information processing systems, medical professionals had to manually input patient information and then create diagnoses and documents based on that information, which was a cumbersome process. It also required a great deal of time and effort to efficiently search for related case information based on individual patient information and obtain appropriate diagnostic results. Furthermore, the system lacked the functionality to automatically generate and edit documents based on diagnostic results, increasing the burden on medical professionals. In addition, improving the accuracy of machine learning models also required manual effort, making the automation and accuracy improvement of the entire system a challenge.

[1275] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1276] In this invention, the server includes an information input means for inputting patient information, a keyword generation means for generating keywords based on the input patient information, a data acquisition means for acquiring case information from an external database using the generated keywords, a probability estimation means for estimating the probability of disease based on the acquired case information, an information display means for displaying the estimated probability information, a document generation means for automatically generating a medical certificate or prescription based on the estimation result, a document editing means for displaying the document generated by the document generation means in a format that can be edited by the user, and a learning means for improving the accuracy of the system using stored data. This reduces the burden on medical professionals and enables efficient and accurate diagnosis and document creation.

[1277] "Patient information" is a general term for a patient's basic personal information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[1278] "Information input means" refers to devices and software that allow a user to input patient information, and includes terminals and special input software.

[1279] "Keyword generation means" refers to a means for automatically extracting key keywords from related symptoms and test results based on input patient information.

[1280] "Data acquisition means" refers to a means for searching and acquiring case information from external databases or the Internet using the generated keywords.

[1281] "Probability estimation means" refers to a means of estimating the probability of disease occurrence based on acquired case information using statistical analysis and machine learning techniques.

[1282] "Information display means" refers to a device or software for visually displaying estimated probability information to a user, including display in list or graph format.

[1283] "Document generation means" refers to software or algorithms for automatically generating medical certificates and prescriptions based on the inference results.

[1284] The "document editing means" refers to a means for providing an interface that allows a user to modify or add to a generated document.

[1285] "Learning means" refers to the means of training a machine learning model using stored patient information and diagnostic results to improve the accuracy of the system.

[1286] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and, based on that information, assists in disease probability estimation and document creation. The system includes an information input means for inputting patient information, a keyword generation means for generating keywords, a data acquisition means for searching case information, a probability estimation means for analyzing the acquired information, an information display means for displaying the analysis results, a document generation means for automatically generating documents, a document editing means for editing the generated documents, and a learning means for improving the accuracy of the system.

[1287] Hardware and Software Configuration

[1288] Information input method

[1289] User:

[1290] Users (healthcare workers) log in to a dedicated terminal (PC, tablet, etc.), which is equipped with an easy-to-use GUI (graphical user interface), and enter basic patient information (name, age, gender, etc.) and specific symptoms (e.g., headache, fever, cough).

[1291] Device:

[1292] The device has a software module that converts the information into a standard format, such as JSON or XML, and then the converted data is encrypted and securely sent to the server.

[1293] Probability estimation means and information display means

[1294] server:

[1295] The server stores the received patient information in a database and performs an initial consistency check. It then uses natural language processing (NLP) technology to extract relevant keywords from the data. The generated keywords are then used to search for related cases and literature using medical database APIs and search engines. After filtering, the most reliable information is used to estimate the probability of disease using statistical analysis and machine learning algorithms.

[1296] Device:

[1297] The terminal visually displays the estimation results, which can be displayed in list or graph format for the user to review.

[1298] Document creation and editing tools

[1299] server:

[1300] The server prepares templates for medical certificates and prescriptions based on the estimation results, and automatically generates draft documents by inserting the patient's individual information. These templates are predefined.

[1301] Device:

[1302] The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections or additions. The user can then check, correct, and finalize the document.

[1303] Learning tools

[1304] server:

[1305] The server uses the stored data to train the machine learning model, which is periodically retrained with new data sets to improve the system's accuracy. This retraining process uses advanced machine learning techniques such as deep learning.

[1306] Specific examples

[1307] For example, if a patient complains of "headache, fever, cough," the following process occurs:

[1308] User:

[1309] Symptoms are entered into the terminal, which converts them into a standard format and sends them to the server via encrypted communication.

[1310] server:

[1311] It receives information, generates keywords such as "headache," "fever," and "cough," and searches and filters related cases and literature. For example, it estimates the probability of influenza at 40%, cold at 30%, and pneumonia at 20%.

[1312] Device:

[1313] The system displays the estimated results and automatically generates the necessary medical certificates and prescriptions, which the user can then review and finalize.

[1314] Prompt Sentence Examples

[1315] "The patient's symptoms are headache, fever, and cough. Past medical history includes allergic rhinitis, and current test results show an elevated white blood cell count. Based on this information, generate search keywords and related case information, and calculate the estimated probability of the most relevant disease. Additionally, create a draft medical certificate based on the results."

[1316] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1317] Step 1:

[1318] The user enters information

[1319] Input: Patient basic information (name, age, gender, etc.), specific symptoms (e.g. headache, fever, cough), past medical history, and current test results.

[1320] How it works: The user (healthcare worker) logs in to a dedicated device (PC, tablet, etc.) and enters patient information. The user enters information using an easy-to-use GUI.

[1321] Output: The entered patient information is temporarily saved in the terminal's memory.

[1322] Step 2:

[1323] The device prepares the data

[1324] Input: Patient information entered by the user.

[1325] How it works: The device converts the input information into a standard format (e.g., JSON or XML) to facilitate further processing. The converted data is then encrypted and security protocols are applied to prevent external access.

[1326] Output: The encrypted data converted into a standard format.

[1327] Step 3:

[1328] The device sends the data to the server

[1329] Input: Encrypted data converted to a standard format.

[1330] Operation: The device sends encrypted data to the server. The communication protocol is SSL / TLS, and encryption is used to prevent unauthorized access or tampering of data. After the transmission is complete, the device notifies the user that the transmission was successful.

[1331] Output: Data sent to the server and notification to the user that the data has been sent.

[1332] Step 4:

[1333] The server receives and stores the data

[1334] Input: Encrypted data sent from the terminal.

[1335] Operation: The server decrypts the received data and stores it in the database. A relational database is used as the database system. Before storing it in the database, the data is checked for consistency, and an error log is recorded if there is an inconsistency.

[1336] Output: Patient information stored in database and error log (if required).

[1337] Step 5:

[1338] The server generates the keywords

[1339] Input: Patient information stored in the database.

[1340] How it works: The server uses natural language processing (NLP) techniques to extract key keywords from the patient information entered. For example, if the patient information contains "headache," "fever," and "cough," those keywords are generated.

[1341] Output: The generated keywords.

[1342] Step 6:

[1343] The server searches for case information

[1344] Input: Generated keywords.

[1345] How it works: The server uses the generated keywords to search for case reports and literature on the Internet using medical database APIs and search engines. It queries specific databases (e.g., medical research databases or academic paper databases) to obtain relevant information. The obtained data is then filtered to select information that is highly reliable and closely matches the keywords.

[1346] Output: Filtered case information.

[1347] Step 7:

[1348] The server estimates the probability of the disease

[1349] Input: Filtered case information.

[1350] How it works: The server uses statistical analysis and machine learning algorithms to estimate disease probabilities based on filtered case information, such as "the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%."

[1351] Output: Estimated probability information.

[1352] Step 8:

[1353] The server sends the estimation results to the device.

[1354] Input: Estimated probability information.

[1355] How it works: The server converts the probability information into JSON format and sends it to the device. The transmission process is secure using the SSL / TLS protocol.

[1356] Output: Probability information sent to the terminal.

[1357] Step 9:

[1358] The device displays the estimated results.

[1359] Input: Probability information sent to the terminal.

[1360] Operation: The device visually displays the received probability information in a list or graph format designed to be easily understood by the user.

[1361] Output: Probability information displayed to the user.

[1362] Step 10:

[1363] The server automatically generates the document

[1364] Input: Estimated probability information and basic patient information.

[1365] How it works: The server prepares templates for medical certificates and prescriptions based on the inference results, embeds the patient's individual information, and automatically generates draft documents using a template engine.

[1366] Output: Auto-generated draft document.

[1367] Step 11:

[1368] The device edits and displays the document.

[1369] Input: Auto-generated draft document.

[1370] Operation: The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections and additions. Once editing is complete, the user performs a final check and finalizes the document.

[1371] Output: Final revised and verified document.

[1372] Step 12:

[1373] The server stores the data and updates the learning model.

[1374] Input: Finalized documentation and diagnostic results.

[1375] How it works: The server stores the final diagnosis and treatment progress in a database. The stored data is used to periodically retrain the machine learning model to improve the accuracy of the system. The machine learning algorithm includes adaptive learning to accommodate increasing data.

[1376] Output: Updated machine learning model and saved diagnostic data.

[1377] (Application example 1)

[1378] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1379] In modern factories, maintenance and inspection of each machine is still often done manually. This relies on the knowledge and experience of the maintenance personnel, which can delay early detection of abnormalities and optimal maintenance responses. Furthermore, updating maintenance histories and manuals is time-consuming, which tends to lower the efficiency of the entire factory. Furthermore, identifying the cause of an abnormality and selecting a countermeasure can take time, so an efficient system to solve these problems is needed.

[1380] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1381] In this invention, the server includes input means for inputting patient information, means for generating keywords based on the input patient information, search means for retrieving case information from an external database using the generated keywords, analysis means for estimating the probability of disease based on the retrieved case information, means for displaying the estimated probability information, input means for inputting each device information, means for generating keywords based on the input device information, search means for retrieving maintenance response actions from an external database using the generated keywords, and means for applying the retrieved maintenance response actions to machine maintenance and repair. This makes it possible to efficiently and quickly identify machine abnormalities and propose optimal maintenance responses.

[1382] "Patient information" refers to basic patient data (such as name, age, sex, past medical history, specific symptoms, and test results) entered by medical professionals.

[1383] "Input means" refers to equipment or devices (PCs, tablets, smartphones, etc.) used to collect patient information and information on machinery and equipment and input it as data.

[1384] "Keyword generation means" refers to the technology or algorithm used to extract key keywords from input information.

[1385] A "search means" is a system or software that uses the generated keywords to retrieve related information from an external database.

[1386] "Analysis means" refers to a device or program that has the function of estimating and displaying specific results using statistical analysis or machine learning based on acquired information.

[1387] "Display means" refers to a display or monitor for visually displaying analysis results, estimated probability information, maintenance response actions, etc.

[1388] "Each device information" refers to data on the status and abnormalities of various machinery and equipment within the factory (such as years of operation, error logs, and current operating status).

[1389] "Maintenance response actions" refer to specific procedures and tasks for resolving abnormalities or breakdowns in machinery and equipment.

[1390] "Document generation means" refers to software or a system for automatically generating documents such as medical certificates, prescriptions, and maintenance manuals based on acquired information.

[1391] "Learning" is the process of using stored data to train machine learning models to improve the system's analytical accuracy and predictive capabilities.

[1392] This invention relates to a "smart maintenance system" for improving the efficiency of maintenance and inspection work for factory robots and machinery equipment. In particular, it is a system for automating a series of processes including input of equipment information, keyword generation, related information search, analysis, display, document generation, and learning.

[1393] 1. Entering device information and initial processing

[1394] User

[1395] Users (maintenance personnel) use dedicated devices (smartphones, tablets, etc.) to input information about each piece of machinery and equipment. For example, they input data such as the machine's name, years of operation, current status, and error logs. This information plays an important role in subsequent analysis and maintenance measures.

[1396] Terminal

[1397] The terminal converts the input data into a standard format and transmits it to the server using secure encrypted communications. After transmitting the data, the terminal notifies the user that the transmission has been completed.

[1398] 2. Data Acquisition and Analysis

[1399] server

[1400] The server stores the machine equipment data sent from the terminal in a database. At the same time, it checks the data for consistency. Next, it uses natural language processing (NLP) to generate keywords from the input information. For example, it extracts key keywords from error logs, such as "loud operating noise" or "vibration felt."

[1401] Based on the generated keywords, the server searches an external database for relevant maintenance actions, filtering the results based on criteria such as reliable data sources and keyword match.

[1402] 3. Estimation and provision of information on maintenance response actions

[1403] server

[1404] The server analyzes the filtered information and predicts the most appropriate maintenance action. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." The predicted maintenance action is sent to the terminal in JSON format.

[1405] Terminal

[1406] The terminal visually displays the received maintenance response actions. Possible display formats include lists, graphs, and text. In addition, documents such as maintenance manuals and maintenance histories are automatically generated based on the estimation results. The generated documents are displayed in a format that can be checked and edited by the user.

[1407] 4. Document Creation and Editing

[1408] server

[1409] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This document generation is performed based on predefined templates.

[1410] Terminal

[1411] The terminal displays the generated draft document to the user and provides an editing interface that allows the user to make corrections or additions as necessary. After the user checks and corrects the document, it is finalized.

[1412] 5. Data accumulation and accuracy improvement

[1413] server

[1414] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved.

[1415] Specific examples

[1416] The maintenance technician enters the status information of the "Robot Arm A1" (2 years of operation, operating noise, vibration detection) into the terminal. The input data is converted to JSON format and sent to the maintenance server.

[1417] The server analyzes the error log information sent and extracts key keywords ("operating sound" and "vibration") using TF-IDF. Maintenance response actions are searched and filtered based on the obtained keywords.

[1418] Based on the extracted keywords, the most appropriate maintenance actions ("Check the cooling system and clean the filters" and "Check for loose parts and lubricate the bearings") are displayed.

[1419] Prompt Sentence Examples

[1420] "You input the machine's condition, for example, 'it's loud when it's running' or 'I feel vibrations.' Then, based on this information, we suggest the best maintenance action."

[1421] As described above, the present invention provides a system that improves the efficiency of maintenance management within a factory, reduces the burden on maintenance personnel, and enables early detection of machine abnormalities.

[1422] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1423] Step 1:

[1424] Users input information about each piece of machinery and equipment using a dedicated device (smartphone, tablet, etc.). The input includes the machine's name, years of operation, current status, error log, etc. This allows data input (machine information) from the user to be obtained. This data is the basic data required for subsequent analysis.

[1425] Step 2:

[1426] The terminal converts the machine information entered by the user into a standard format, for example, JSON format, and sends it to the server using encrypted communication. This allows data to be sent from the terminal to the server.

[1427] Step 3:

[1428] The server receives the machine information sent from the terminal and stores it in a database. Before storing it, it checks the data for consistency and ensures that there are no abnormalities. This ensures the reliability of the data.

[1429] Step 4:

[1430] The server performs natural language processing (NLP) based on the received machine information to extract key keywords. For example, keywords such as "loud operating noise" and "vibration felt" are generated from error logs using TF-IDF. This provides basic keywords for analysis.

[1431] Step 5:

[1432] The server queries an external database (such as a maintenance manual database) based on the generated keywords. It extracts appropriate maintenance response actions and collects information on them. Search results are filtered based on reliable data sources, keyword matching, etc.

[1433] Step 6:

[1434] The server performs statistical analysis on the acquired maintenance response actions and estimates specific maintenance measures. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." This allows it to estimate the optimal maintenance response action.

[1435] Step 7:

[1436] The server converts the estimated maintenance action into JSON format and sends it to the terminal, which then transmits the analysis results to the terminal.

[1437] Step 8:

[1438] The terminal visually displays the received maintenance action and notifies the user. The display format can be a list, graph, text, etc. This allows the user to confirm the maintenance action.

[1439] Step 9:

[1440] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This generates documents for maintenance actions.

[1441] Step 10:

[1442] The terminal displays the generated draft document to the user and provides an editing interface that allows corrections and additions as necessary. After the user checks and corrects the document, they finalize it. This finalizes the maintenance document.

[1443] Step 11:

[1444] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved, thereby improving overall system performance.

[1445] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1446] The present invention relates to an information provision system that efficiently inputs patient information and, based on that information, estimates disease probability, creates documents, and recognizes user emotions. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means. Furthermore, it incorporates an emotion engine and has the function of customizing the method of providing diagnosis results and advice based on the user's emotional state.

[1447] 1. Entering patient information and initial processing

[1448] User

[1449] The user (healthcare worker) logs in to a dedicated terminal and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly into the terminal.

[1450] Terminal

[1451] The terminal converts the input information into a standard format and transmits it to the server using encrypted communications. Once transmission is complete, the terminal notifies the user.

[1452] 2. Data Acquisition and Analysis

[1453] server

[1454] The server stores the patient information sent from the terminal in a database, checks the data consistency, and sends a supplemental request to the terminal if any information is missing.

[1455] server

[1456] The server analyzes stored patient information and generates keywords for related symptoms, using natural language processing (NLP) to extract key keywords from symptoms and test results.

[1457] server

[1458] The server searches online literature databases and case report sites based on the generated keywords, and retrieves relevant case information using medical database APIs and search engine queries.

[1459] server

[1460] The server performs filtering to extract highly relevant information from the acquired case information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[1461] server

[1462] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, it may estimate "40% probability of influenza," "30% probability of cold," or "20% probability of pneumonia." These estimation results are converted into JSON format and sent to the device.

[1463] 3. Information provision and document generation

[1464] Terminal

[1465] The device visually displays the received information so that the user can check it. It can use display formats such as lists and graphs. Furthermore, it automatically generates medical certificates, prescriptions, referral letters, etc. based on the inference results.

[1466] server

[1467] Based on the estimation results, the server prepares templates for medical certificates and prescriptions, inserts the patient's individual information, and generates draft documents.

[1468] Terminal

[1469] The terminal provides the generated draft document in an editable format, allowing the user to check and modify the draft document and make any necessary additions.

[1470] 4. Emotion Recognition and Response

[1471] Terminal

[1472] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[1473] server

[1474] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[1475] server

[1476] Based on the assessment, the server adjusts the presentation of the diagnostic results to provide information in a format that is most acceptable to the user, for example, using calmer, more reassuring language for a user in a stressed state.

[1477] Terminal

[1478] The device uses an emotion engine to provide users with customized medical advice, for example, providing detailed instructions on specific steps to take or the need for additional testing if the user is worried.

[1479] 5. Data accumulation and learning

[1480] server

[1481] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[1482] As a concrete example, if a patient complains of "headache, fever, and cough," the user enters this information into the terminal, and the server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, the terminal will display this information and further analyze the user's emotional state to provide information in the most optimal form. Medical certificates and prescriptions are also automatically generated, with the user making the final confirmation.

[1483] This system can provide information quickly and accurately, reducing the burden on medical professionals. Furthermore, emotion recognition can be used to present information in a way that is easy for users to accept, improving the quality of medical care.

[1484] The processing flow will be explained below.

[1485] Step 1: Enter patient information

[1486] User

[1487] The user (medical worker) logs in to a dedicated terminal.

[1488] The user enters the patient's basic information (name, age, gender, etc.), symptoms (headache, fever, cough, etc.), past medical history, and current test results.

[1489] Step 2: Sending data

[1490] Terminal

[1491] The terminal converts the entered patient information into a standard format.

[1492] The terminal transmits the patient information to the server using encrypted communication.

[1493] The terminal notifies the user that the transmission is complete.

[1494] Step 3: Save data and check its integrity

[1495] server

[1496] The server stores the received patient information in a database.

[1497] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[1498] Step 4: Generate keywords

[1499] server

[1500] The server analyzes stored patient information and automatically generates keywords for relevant symptoms.

[1501] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[1502] Step 5: Internet Search

[1503] server

[1504] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[1505] Use medical database APIs or search engine queries to retrieve relevant case information.

[1506] Step 6: Filtering case information

[1507] server

[1508] The server performs filtering to extract highly relevant information from the acquired case information.

[1509] Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[1510] Step 7: Estimate the probability of disease

[1511] server

[1512] The server performs statistical analysis based on the filtered case information and estimates the probability of each disease.

[1513] For example, estimates might say, "The probability of influenza is 40%," "The probability of a cold is 30%," and "The probability of pneumonia is 20%."

[1514] Step 8: Submit your information

[1515] server

[1516] The server converts the estimated probability information into JSON format and sends it to the terminal.

[1517] Step 9: Viewing Information

[1518] Terminal

[1519] The device visually displays the information it receives, for example in the form of a list or graph.

[1520] The user reviews and evaluates the displayed information.

[1521] Step 10: Emotion Recognition

[1522] Terminal

[1523] The device uses an emotion engine to analyze the user's facial expressions, tone of voice, and input content while operating the device, and recognizes the user's emotional state.

[1524] Step 11: Assess and regulate your emotional state

[1525] server

[1526] The server evaluates the user's emotional state (stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[1527] The server adjusts the way the information is presented based on the evaluation results, and provides the information in a format that is easy for the user to accept.

[1528] Step 12: Customize your information offering

[1529] Terminal

[1530] The terminal uses the emotion engine to present customized medical advice to the user.

[1531] For example, a user in a good emotional state may be provided with concise and efficient information, while a user experiencing stress may be provided with detailed and reassuring information.

[1532] Step 13: Document Generation

[1533] server

[1534] The server prepares templates for a diagnosis, prescription, and referral letter based on the estimation results and the user's emotional state.

[1535] Insert the patient's individual information into the template and automatically generate a draft document.

[1536] Step 14: Edit and review the document

[1537] Terminal

[1538] The terminal displays the generated draft document to the user and provides it in an editable format.

[1539] The user checks and corrects the draft document and makes any necessary additions.

[1540] Step 15: Save your data

[1541] server

[1542] The server stores the final diagnosis results and treatment progress in a database.

[1543] Step 16: Training the system and improving accuracy

[1544] server

[1545] The server uses the stored data to train the machine learning model.

[1546] The model is periodically retrained with new datasets to improve the accuracy of the system.

[1547] Example 2

[1548] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1549] In modern medical settings, entering and analyzing patient information, estimating diagnostic probabilities, and determining optimal treatment plans requires a great deal of time and effort. Furthermore, there is often a lack of information provided that takes into account the patient's emotions and mental state, which increases the burden on medical professionals and increases patient anxiety. To efficiently resolve these issues and improve the quality of medical care, a comprehensive and flexible information provision system is required.

[1550] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1551] In this invention, the server includes means for inputting patient information, means for converting the input patient information into a standard format and transmitting it, means for saving the transmitted patient information, means for generating keywords based on the saved patient information, means for retrieving case information from an external database using the generated keywords, means for filtering the retrieved case information based on relevance, means for estimating disease probability based on the filtered case information, means for displaying the estimated probability information, and means for analyzing user emotions and adjusting the information presentation method, thereby enabling efficient input and analysis of patient information, accurate disease probability estimation, and information provision that takes patient emotions into consideration.

[1552] "Patient information" is a general term for basic patient information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[1553] "Means for inputting" refers to a device or interface that allows a user to input patient information.

[1554] "Means for converting into a standard format and transmitting" refers to the function for converting input patient information into a specified format and transmitting it securely to a server using an encrypted communication protocol.

[1555] "Means for storing" refers to a function for storing transmitted patient information in a database.

[1556] "Means for generating keywords" refers to technology that analyzes stored patient information and extracts keywords based on key symptoms and test results.

[1557] "Means for obtaining case information" refers to a function for obtaining related case information from external databases or case report sites using the generated keywords.

[1558] "Filtering means" refers to a function for sorting acquired case information based on relevance and extracting only highly reliable information.

[1559] "Means for estimating the probability of disease" refers to technology for performing statistical analysis based on filtered case information to calculate the probability of occurrence of a specific disease.

[1560] "Means for displaying probability information" refers to a display interface that allows a user to visually confirm the estimated disease occurrence probability.

[1561] "Means for analyzing emotions and adjusting the way information is presented" refers to a function that analyzes the user's emotional state, such as facial expressions and tone of voice, and customizes the way information is presented based on the results.

[1562] "Means for automatically generating documents" refers to a function for automatically generating medical documents such as medical certificates, prescriptions, and referral letters.

[1563] "Means for displaying in an editable format" refers to an interface for displaying the generated medical document in a format that can be checked and corrected by the user.

[1564] "Means for training machine learning models to improve the accuracy of the system" refers to technology that uses stored data on past diagnostic results and treatment progress to train machine learning models and improve the diagnostic accuracy of the system.

[1565] The present invention relates to an information providing system that efficiently inputs patient information and, based on the input information, estimates the probability of disease, creates documents, and recognizes user emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[1566] System configuration and basic operations

[1567] User

[1568] The user (healthcare worker) logs in to a dedicated terminal. The terminal provides an interface for entering basic patient information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly by hand. Once the user has finished entering the patient information, the terminal converts this information into a standard format and sends it to the server using an encrypted communication protocol (e.g., HTTPS).

[1569] Terminal

[1570] The terminal converts the received patient information into a standard JSON format and sends the information to the server using the HTTPS protocol. Once the transmission is complete, a notification of the completion is displayed to the user.

[1571] server

[1572] The server stores the patient information sent from the device in a database. It checks the consistency of the stored information and sends a supplemental request to the device if any information is missing. After consistency is confirmed, the server uses natural language processing (NLP) technology (e.g., spaCy or NLTK) to generate keywords from the patient information based on key symptoms and test results. Using the generated keywords, the server retrieves relevant case information from external literature databases and case report sites via medical database APIs (e.g., PubMed, ClinicalTrials.gov) or search engines.

[1573] The server filters the acquired case information and selects only the most relevant information. Based on the filtered information, the server performs statistical analysis to estimate the probability of a specific disease occurring. The estimated results (e.g., "40% probability of influenza," "30% probability of cold," "20% probability of pneumonia") are converted to JSON format and sent back to the terminal.

[1574] Linking devices and servers

[1575] The terminal visually displays the received probability estimation results, using formats such as lists and graphs. Furthermore, based on the information sent from the server, the terminal automatically generates medical documents such as medical certificates, prescriptions, and referral letters. The generated documents are provided in an editable format that the user can review and modify.

[1576] Emotion recognition function

[1577] When the user operates the device, the device uses an emotion engine to recognize the user's emotional state, such as facial expressions and tone of voice. The server evaluates the user's emotional state based on the analysis results of the emotion engine and adjusts the way the diagnosis results are presented. For example, if the user is anxious, the system will display information in calm language and provide detailed instructions on how to deal with the situation.

[1578] Learning Function

[1579] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model (e.g., neural network) to improve the accuracy of the system.

[1580] Specific examples

[1581] For example, if a patient complains of headache, fever, and cough, the user enters this information into the device. The server uses natural language processing to automatically generate keywords such as "headache," "fever," and "cough," and searches the case database. The search results include an estimate of a 40% probability of influenza, a 30% probability of a cold, and a 20% probability of pneumonia. These estimates are sent to the device in JSON format, where they are displayed visually in graphs. The emotion engine analyzes the user's facial expressions and provides customized advice, such as displaying the diagnosis results in calm language. Furthermore, a medical certificate and prescription are automatically generated, and the user makes the final confirmation.

[1582] Example prompts for generative AI models

[1583] "A patient complains of headache, fever, and cough. Their only past medical history is allergic rhinitis. Estimate the probability of influenza and visually display the results. Also, if the user is stressed, please respond by providing information in a calmer tone."

[1584] This system enables prompt and accurate information provision, reducing the burden on medical professionals. In addition, the emotion recognition function allows information to be provided in a format that is easy for users to accept, improving the quality of medical care.

[1585] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1586] Step 1:

[1587] User login and patient information entry

[1588] User

[1589] The user logs in to a dedicated terminal. They access the login screen, enter their user ID and password, and click the login button. If login is successful, they are redirected to the patient information input screen. The user enters the patient's basic information (name, age, sex), specific symptoms (headache, fever, cough, etc.), past medical history, and current test results in the form on the screen, and clicks the "Submit" button.

[1590] Input and Output

[1591] Input: Patient information entered by the user into the terminal

[1592] Output: The input information is converted into a standard format.

[1593] Step 2:

[1594] Converting and sending input information

[1595] Terminal

[1596] The terminal converts the received patient information into a standard format (e.g., JSON format) and sends it to the server using an encrypted communication protocol (e.g., HTTPS). Once the transmission is complete, the terminal displays a pop-up message stating, "Information transmission completed."

[1597] Input and Output

[1598] Input: User-entered patient information

[1599] Output: Information converted to standard format and notification of completion of transmission

[1600] Step 3:

[1601] Server information storage and supplementary requests

[1602] server

[1603] The server receives the patient information sent from the terminal and stores it in the database. After storing it, it checks the consistency of the data and sends a supplementary request to the terminal if necessary information is missing.

[1604] Input and Output

[1605] Input: Patient information submitted in a standard format

[1606] Output: Stored information and supplementary requests after consistency check

[1607] Step 4:

[1608] Generate keywords for patient information

[1609] server

[1610] The server analyzes the stored patient information and generates keywords based on relevant symptoms and test results using natural language processing (NLP) techniques (e.g., spaCy and NLTK).

[1611] Input and Output

[1612] Input: Saved patient information

[1613] Output: Generated keyword list

[1614] Step 5:

[1615] Case database search

[1616] server

[1617] The server searches external literature databases and case report sites based on the generated keywords, using medical database APIs and search engine queries.

[1618] Input and Output

[1619] Input: Generated keyword list

[1620] Output: Retrieved case information

[1621] Step 6:

[1622] Case information filtering

[1623] server

[1624] The server then performs a filtering process based on the acquired case information to extract highly relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[1625] Input and Output

[1626] Input: Retrieved case information

[1627] Output: Filtered case information list

[1628] Step 7:

[1629] Disease probability estimation and result transmission

[1630] server

[1631] The server performs statistical analysis based on the filtered case information to estimate the probability of candidate diseases. The estimated results are converted into JSON format and sent to the device.

[1632] Input and Output

[1633] Input: Filtered case information

[1634] Output: Estimated disease probability information

[1635] Step 8:

[1636] Displaying diagnostic results and generating documentation

[1637] Terminal

[1638] The device visually displays the received information for the user to review. Display formats include lists and graphs. Furthermore, medical documents such as medical certificates, prescriptions, and referral letters are automatically generated based on the inference results. The generated documents are provided in an editable format, allowing the user to review and correct them.

[1639] Input and Output

[1640] Input: Estimated disease probability information

[1641] Output: Visually displayed information and generated medical documents

[1642] Step 9:

[1643] Emotion recognition and information presentation adjustment

[1644] Terminal

[1645] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[1646] server

[1647] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine. Based on the evaluation results, it adjusts the way the diagnosis results are presented to provide information in a format that is most acceptable to the user. For example, for a user in a stressed state, it uses calmer, more reassuring language.

[1648] Terminal

[1649] The device uses an emotion engine to provide users with customized medical advice, detailing specific steps to take or the need for additional testing if the user is concerned.

[1650] Input and Output

[1651] Input: User's emotional state

[1652] Output: customized diagnostic results

[1653] Step 10:

[1654] Data accumulation and learning

[1655] server

[1656] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[1657] Input and Output

[1658] Input: Final diagnosis and treatment progress

[1659] Output: Updated machine learning model

[1660] Specific actions

[1661] In this way, through a series of processes, patient information is efficiently input and analyzed, appropriate disease probability is estimated, and optimal treatment plans are determined.In addition, by presenting information that takes the user's emotions into consideration, the burden on medical professionals is reduced and patients feel more at ease.

[1662] (Application example 2)

[1663] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1664] Conventional health management systems for factory workers have difficulty efficiently integrating multiple health data sets and quickly and accurately assessing health risks. Furthermore, they do not provide appropriate advice based on the worker's emotional state, resulting in decreased productivity and the risk of workers' health deteriorating. These issues needed to be resolved.

[1665] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1666] In this invention, the server includes an input means for inputting the health data of workers, a means for generating keywords based on the input health data, a search means for retrieving case information from an external database using the generated keywords, an analysis means for estimating the probability of a health risk based on the retrieved case information, a means for displaying the estimated probability information, an emotion recognition means for recognizing the emotional state of the user, a means for customizing the information provision method based on the recognized emotional state, a document generation means for automatically generating documents such as medical certificates and prescriptions, a means for displaying the generated documents in an editable format, and a means for estimating health risks based on the health data and suggesting preventive measures. This makes it possible to efficiently monitor and evaluate the health status of factory workers and provide appropriate advice according to their emotional state.

[1667] "Patient information" refers to basic information about an individual patient collected by a medical professional, including, specifically, data such as the patient's name, age, gender, symptoms, medical history, and current test results.

[1668] "Input means" refers to an interface or device that allows a user to input patient information into a terminal.

[1669] The "keyword generating means" is a means for extracting major symptoms and related information from input patient information and generating search keywords.

[1670] The "search means" is a means for retrieving related case information from an external database using the generated keywords.

[1671] The "analysis means" is a means for estimating the probability of a particular disease or health risk based on the acquired case information.

[1672] The "display means" is a means for visually displaying estimated probability information and diagnostic results in a format that can be confirmed by the user.

[1673] The "emotion recognition means" is a means for analyzing the user's facial expressions and tone of voice to recognize the user's emotional state.

[1674] The "information presentation method customization means" is a means for optimizing the information presentation method based on the recognized emotional state.

[1675] The "document generation means" is a means for automatically creating documents such as medical certificates, prescriptions, and referral letters.

[1676] "Means for displaying in an editable format" refers to means for providing the generated document in a format that allows the user to check and modify it.

[1677] "Healthcare data" refers to data that indicates the health status of factory workers, specifically including blood pressure, heart rate, symptoms, medical history, etc.

[1678] "Means for estimating health risks" are means for assessing the need for preventive measures and countermeasures based on health data and for calculating the probability of risk.

[1679] A "means for proposing preventive measures" is a means for proposing specific preventive measures or countermeasures based on estimated health risks.

[1680] This invention is applied to a health management system for factory workers. Its basic components include an input means for inputting health data, an analysis means for standardizing and analyzing the data, a display means for displaying and editing estimation results and documents, and an emotion recognition means for recognizing emotional states. It also includes a means for assessing health risks and suggesting individual preventive measures.

[1681] 1. Health data entry and initial processing

[1682] User

[1683] Users (factory workers or managers) use dedicated input devices (e.g., PCs or tablets) to enter health data into the system. This data includes blood pressure, heart rate, symptoms, medical history, etc. After completing the data entry, the user sends the data from their device to the server. At this time, the data is converted into a standard format and sent to the server using encrypted communication.

[1684] 2. Data Acquisition and Analysis

[1685] server

[1686] The server receives the health data sent from the device and stores it in a database. Natural language processing (NLP) is used to generate keywords from the stored data, and relevant case information is searched for in external databases. Search methods include medical database APIs and search engine queries.

[1687] server

[1688] The server filters the acquired case information and estimates health risks based on reliable information. For example, it may estimate "40% probability of overwork," "30% probability of high blood pressure," or "20% probability of myocardial infarction." The estimated results are converted into JSON format and sent to the device.

[1689] 3. Information provision and document generation

[1690] Terminal

[1691] The device visually displays the received information for the user to review, and can use lists and graphs as display formats. It can also automatically generate and display documents such as medical certificates, prescriptions, and referral letters in user-editable formats.

[1692] 4. Emotion Recognition and Response

[1693] Terminal

[1694] When the user operates the terminal, the terminal uses emotion recognition means to recognize the user's emotional state from the user's facial expression, tone of voice, input contents, etc.

[1695] server

[1696] Based on the analysis results of the emotion recognition means, the server evaluates the user's emotional state (e.g., stress, relief, fatigue, etc.) and optimizes the way it provides information. For example, if the user is fatigued, it will provide detailed, specific instructions on how to deal with the situation in a gentle manner.

[1697] Terminal

[1698] The device uses emotion recognition to provide customized advice to the user, detailing specific actions to take or the need for further testing if the user is concerned.

[1699] 5. Data accumulation and learning

[1700] server

[1701] The server stores the final diagnosis results and countermeasure results in a database and uses the stored data to train the machine learning model, thereby improving the accuracy of the system.

[1702] Examples:

[1703] For example, suppose Mr. Tanaka (not his real name) complains of fatigue, headache, and muscle pain during a health checkup. This information is entered into the terminal. Based on this, the server stores the keywords "fatigue," "headache," and "muscle pain," and retrieves case information from an external database. As a result, it estimates that there is a "40% chance of overwork," a "30% chance of high blood pressure," and a "20% chance of myocardial infarction," and Mr. Tanaka's emotional state is also analyzed. If Mr. Tanaka is feeling stressed, it gently explains how to deal with the situation.

[1704] Example prompt sentence:

[1705] "Design a system to monitor and analyze the health status of factory workers. Analyze the entered health data (e.g., symptoms, medical history, current test results) and estimate health risks. Additionally, analyze the user's emotional state and provide individually customized advice."

[1706] In this way, by using this system, it is possible to efficiently manage the health of factory workers and provide appropriate advice according to the emotional state of each individual worker.

[1707] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1708] Step 1:

[1709] The user enters health data.

[1710] Input: The user inputs health data such as blood pressure, heart rate, symptoms, and medical history into the device.

[1711] Processing: The entered data is converted into a standard format and sent to the server using encrypted communications.

[1712] Output: A confirmation of receipt of the data sent is displayed on the terminal.

[1713] Step 2:

[1714] The server receives and stores the data.

[1715] Input: Health data sent from the device in a standard format.

[1716] Processing: Stores the received data in a database, checks the data consistency, and sends a supplemental request to the terminal if there is missing information.

[1717] Output: The saved data. If additional information is required, the request is notified to the terminal.

[1718] Step 3:

[1719] The server generates the keywords.

[1720] Input: Stored health data.

[1721] Processing: Using natural language processing (NLP), key keywords are automatically extracted. For example, "fatigue," "headache," and "muscle pain" are extracted.

[1722] Output: The generated keywords.

[1723] Step 4:

[1724] The server retrieves case information from an external database.

[1725] Input: Generated keywords.

[1726] Processing: Search for case information based on keywords using medical database APIs and search engine queries.

[1727] Output: Retrieved case information.

[1728] Step 5:

[1729] The server estimates the health risk.

[1730] Input: Retrieved case information.

[1731] Processing: Filtering is performed and statistical analysis is carried out based on highly reliable information. For example, estimates such as "40% probability of overwork," "30% probability of high blood pressure," and "20% probability of myocardial infarction" are made.

[1732] Output: Health risk probability information (JSON format).

[1733] Step 6:

[1734] The device displays the estimated results.

[1735] Input: Probability information of health risks.

[1736] Processing: Visually displaying received information in the form of lists and graphs, providing information in a format that is easy for users to understand.

[1737] Output: Visually displayed estimation results.

[1738] Step 7:

[1739] The device recognizes the emotional state.

[1740] Input: The user's facial expressions, tone of voice, and input content when operating the device.

[1741] Processing: Emotion recognition means are used to perform natural language processing and facial expression analysis to assess the user's emotional state.

[1742] Output: Perceived emotional state (e.g., stressed, relieved).

[1743] Step 8:

[1744] The server customizes how the information is presented.

[1745] Input: Perceived emotional state.

[1746] Treatment: Based on the emotional state, determine how to present information in a way that is most acceptable to the user. For example, use gentle language and detailed instructions for a stressed user.

[1747] Output: A customized way to provide information.

[1748] Step 9:

[1749] The device will provide customized advice.

[1750] Input: Customized information delivery method.

[1751] Processing: Explains health risks and preventative measures to the user in detail and displays auto-generated documents such as medical certificates and prescriptions in an editable format.

[1752] Output: Customized advice and editable documentation.

[1753] Step 10:

[1754] The server stores the data and trains the machine learning model.

[1755] Input: Final diagnosis results and countermeasure results.

[1756] Processing: The data is stored in a database and used to train a machine learning model, which improves the accuracy of the next analysis.

[1757] Output: A system with improved accuracy.

[1758] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1759] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1760] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1761] [Fourth embodiment]

[1762] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1763] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1764] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1765] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1766] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1767] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1768] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1769] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1770] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1771] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1772] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1773] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1774] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1775] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and assists them in estimating disease probability and creating documents based on that information. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means.

[1776] 1. Entering patient information and initial processing

[1777] User

[1778] The user (healthcare worker) logs in to a dedicated terminal (PC, tablet, etc.) and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, current test results, etc. Test results can be captured using a scanner or entered directly into the terminal.

[1779] Terminal

[1780] The device converts the input information into a standard format and sends it to the server, encrypting the data to prevent it from leaking, and notifying the user that the data has been sent.

[1781] 2. Data Acquisition and Analysis

[1782] server

[1783] The server stores the patient information sent from the device in a database. It also performs a preliminary data consistency check. It automatically generates keywords for related symptoms based on the patient information. This keyword generation is performed using natural language processing (NLP) to extract key keywords from symptoms and test results.

[1784] Based on the generated keywords, the server searches major literature databases and case report sites on the Internet, using medical database APIs and search engine queries. Related cases and literature are retrieved from the search results, and the results are filtered to extract the most relevant information. Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[1785] 3. Probability estimation and information provision

[1786] server

[1787] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, "The probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%." These estimated results are converted into JSON format and sent to the terminal.

[1788] Terminal

[1789] The device visually displays the received information so that the user can check it. Possible display formats include lists and graphs. The device also automatically generates documents such as medical certificates, prescriptions, and referral letters based on the inference results. The generated documents are displayed in an editable format so that the user can check and modify them.

[1790] 4. Document Creation and Editing

[1791] server

[1792] The server prepares templates for medical certificates and prescriptions based on the inference results, and automatically generates draft documents by inserting the patient's individual information. This document generation is based on predefined templates.

[1793] Terminal

[1794] The terminal displays the generated draft document to the user and provides an interface that allows for corrections and additions as necessary. The user then checks and corrects the document and finalizes it based on the results.

[1795] 5. Data accumulation and accuracy improvement

[1796] server

[1797] The server stores the final diagnosis and treatment progress in a database. The stored data is then used to train the machine learning model. The accuracy of the system is improved by periodically retraining the machine learning model with new data sets.

[1798] As a specific example, if a patient complains of "headache, fever, cough," the user enters this information into the terminal and the information is sent to the server. The server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, this information is sent to the terminal for the user to confirm. Furthermore, the necessary medical certificates and prescriptions are automatically generated based on this information, and the user makes the final confirmation.

[1799] In this way, the system can provide information quickly and accurately, reducing the burden on medical professionals.

[1800] The processing flow will be explained below.

[1801] Step 1: Enter patient information

[1802] User

[1803] The user (medical worker) logs in to a dedicated terminal.

[1804] The user enters the patient's basic information (such as name, age, and gender), specific symptoms (e.g., headache, fever, and cough), past medical history, and current test results.

[1805] Step 2: Sending data

[1806] Terminal

[1807] The terminal converts the entered patient information into a standard format.

[1808] The terminal transmits information to the server using encrypted communication.

[1809] The terminal notifies the user that the transmission is complete.

[1810] Step 3: Save data and check its integrity

[1811] server

[1812] The server stores the received patient information in a database.

[1813] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[1814] Step 4: Generate keywords

[1815] server

[1816] The server analyzes the stored patient information and generates keywords for relevant symptoms.

[1817] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[1818] Step 5: Internet Search

[1819] server

[1820] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[1821] Use medical database APIs or search engine queries to retrieve relevant case information.

[1822] Step 6: Filtering case information

[1823] server

[1824] The server performs filtering processing to extract highly relevant information from the acquired case information.

[1825] Filtering criteria include the publication date of the case, reliable data source, and keyword match.

[1826] Step 7: Estimate the probability of disease

[1827] server

[1828] The server performs statistical analysis based on the filtered case information.

[1829] The probability of each disease is estimated, for example, "40% probability of influenza," "30% probability of cold," and "20% probability of pneumonia."

[1830] Step 8: Submit your information

[1831] server

[1832] The server converts the estimation results into JSON format and sends them to the terminal.

[1833] Step 9: Viewing Information

[1834] Terminal

[1835] The device visually displays the received information, for example in the form of a list or graph.

[1836] The user reviews and evaluates the displayed information.

[1837] Step 10: Prepare for document generation

[1838] server

[1839] The server prepares templates for medical certificates, prescriptions, referral letters, etc. based on the estimation results.

[1840] Insert the patient's individual information into the template and generate a draft document.

[1841] Step 11: Editing the Document

[1842] Terminal

[1843] The terminal displays the generated draft document to the user and provides it in an editable format.

[1844] The user checks the draft document and makes any necessary corrections.

[1845] Step 12: Save your data

[1846] server

[1847] The server stores the final diagnosis results and treatment progress in a database.

[1848] Step 13: Improve the accuracy of the system

[1849] server

[1850] The server uses the stored data to train the machine learning model.

[1851] The model is periodically retrained with new datasets to improve the accuracy of the system.

[1852] Example 1

[1853] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1854] In conventional medical information processing systems, medical professionals had to manually input patient information and then create diagnoses and documents based on that information, which was a cumbersome process. It also required a great deal of time and effort to efficiently search for related case information based on individual patient information and obtain appropriate diagnostic results. Furthermore, the system lacked the functionality to automatically generate and edit documents based on diagnostic results, increasing the burden on medical professionals. In addition, improving the accuracy of machine learning models also required manual effort, making the automation and accuracy improvement of the entire system a challenge.

[1855] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1856] In this invention, the server includes an information input means for inputting patient information, a keyword generation means for generating keywords based on the input patient information, a data acquisition means for acquiring case information from an external database using the generated keywords, a probability estimation means for estimating the probability of disease based on the acquired case information, an information display means for displaying the estimated probability information, a document generation means for automatically generating a medical certificate or prescription based on the estimation result, a document editing means for displaying the document generated by the document generation means in a format that can be edited by the user, and a learning means for improving the accuracy of the system using stored data. This reduces the burden on medical professionals and enables efficient and accurate diagnosis and document creation.

[1857] "Patient information" is a general term for a patient's basic personal information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[1858] "Information input means" refers to devices and software that allow a user to input patient information, and includes terminals and special input software.

[1859] "Keyword generation means" refers to a means for automatically extracting key keywords from related symptoms and test results based on input patient information.

[1860] "Data acquisition means" refers to a means for searching and acquiring case information from external databases or the Internet using the generated keywords.

[1861] "Probability estimation means" refers to a means of estimating the probability of disease occurrence based on acquired case information using statistical analysis and machine learning techniques.

[1862] "Information display means" refers to a device or software for visually displaying estimated probability information to a user, including display in list or graph format.

[1863] "Document generation means" refers to software or algorithms for automatically generating medical certificates and prescriptions based on the inference results.

[1864] The "document editing means" refers to a means for providing an interface that allows a user to modify or add to a generated document.

[1865] "Learning means" refers to the means of training a machine learning model using stored patient information and diagnostic results to improve the accuracy of the system.

[1866] The present invention relates to an information provision system that allows medical professionals to efficiently input patient information and, based on that information, assists in disease probability estimation and document creation. The system includes an information input means for inputting patient information, a keyword generation means for generating keywords, a data acquisition means for searching case information, a probability estimation means for analyzing the acquired information, an information display means for displaying the analysis results, a document generation means for automatically generating documents, a document editing means for editing the generated documents, and a learning means for improving the accuracy of the system.

[1867] Hardware and Software Configuration

[1868] Information input method

[1869] User:

[1870] Users (healthcare workers) log in to a dedicated terminal (PC, tablet, etc.), which is equipped with an easy-to-use GUI (graphical user interface), and enter basic patient information (name, age, gender, etc.) and specific symptoms (e.g., headache, fever, cough).

[1871] Device:

[1872] The device has a software module that converts the information into a standard format, such as JSON or XML, and then the converted data is encrypted and securely sent to the server.

[1873] Probability estimation means and information display means

[1874] server:

[1875] The server stores the received patient information in a database and performs an initial consistency check. It then uses natural language processing (NLP) technology to extract relevant keywords from the data. The generated keywords are then used to search for related cases and literature using medical database APIs and search engines. After filtering, the most reliable information is used to estimate the probability of disease using statistical analysis and machine learning algorithms.

[1876] Device:

[1877] The terminal visually displays the estimation results, which can be displayed in list or graph format for the user to review.

[1878] Document creation and editing tools

[1879] server:

[1880] The server prepares templates for medical certificates and prescriptions based on the estimation results, and automatically generates draft documents by inserting the patient's individual information. These templates are predefined.

[1881] Device:

[1882] The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections or additions. The user can then check, correct, and finalize the document.

[1883] Learning tools

[1884] server:

[1885] The server uses the stored data to train the machine learning model, which is periodically retrained with new data sets to improve the system's accuracy. This retraining process uses advanced machine learning techniques such as deep learning.

[1886] Specific examples

[1887] For example, if a patient complains of "headache, fever, cough," the following process occurs:

[1888] User:

[1889] Symptoms are entered into the terminal, which converts them into a standard format and sends them to the server via encrypted communication.

[1890] server:

[1891] It receives information, generates keywords such as "headache," "fever," and "cough," and searches and filters related cases and literature. For example, it estimates the probability of influenza at 40%, cold at 30%, and pneumonia at 20%.

[1892] Device:

[1893] The system displays the estimated results and automatically generates the necessary medical certificates and prescriptions, which the user can then review and finalize.

[1894] Prompt Sentence Examples

[1895] "The patient's symptoms are headache, fever, and cough. Past medical history includes allergic rhinitis, and current test results show an elevated white blood cell count. Based on this information, generate search keywords and related case information, and calculate the estimated probability of the most relevant disease. Additionally, create a draft medical certificate based on the results."

[1896] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1897] Step 1:

[1898] The user enters information

[1899] Input: Patient basic information (name, age, gender, etc.), specific symptoms (e.g. headache, fever, cough), past medical history, and current test results.

[1900] How it works: The user (healthcare worker) logs in to a dedicated device (PC, tablet, etc.) and enters patient information. The user enters information using an easy-to-use GUI.

[1901] Output: The entered patient information is temporarily saved in the terminal's memory.

[1902] Step 2:

[1903] The device prepares the data

[1904] Input: Patient information entered by the user.

[1905] How it works: The device converts the input information into a standard format (e.g., JSON or XML) to facilitate further processing. The converted data is then encrypted and security protocols are applied to prevent external access.

[1906] Output: The encrypted data converted into a standard format.

[1907] Step 3:

[1908] The device sends the data to the server

[1909] Input: Encrypted data converted to a standard format.

[1910] Operation: The device sends encrypted data to the server. The communication protocol is SSL / TLS, and encryption is used to prevent unauthorized access or tampering of data. After the transmission is complete, the device notifies the user that the transmission was successful.

[1911] Output: Data sent to the server and notification to the user that the data has been sent.

[1912] Step 4:

[1913] The server receives and stores the data

[1914] Input: Encrypted data sent from the terminal.

[1915] Operation: The server decrypts the received data and stores it in the database. A relational database is used as the database system. Before storing it in the database, the data is checked for consistency, and an error log is recorded if there is an inconsistency.

[1916] Output: Patient information stored in database and error log (if required).

[1917] Step 5:

[1918] The server generates the keywords

[1919] Input: Patient information stored in the database.

[1920] How it works: The server uses natural language processing (NLP) techniques to extract key keywords from the patient information entered. For example, if the patient information contains "headache," "fever," and "cough," those keywords are generated.

[1921] Output: The generated keywords.

[1922] Step 6:

[1923] The server searches for case information

[1924] Input: Generated keywords.

[1925] How it works: The server uses the generated keywords to search for case reports and literature on the Internet using medical database APIs and search engines. It queries specific databases (e.g., medical research databases or academic paper databases) to obtain relevant information. The obtained data is then filtered to select information that is highly reliable and closely matches the keywords.

[1926] Output: Filtered case information.

[1927] Step 7:

[1928] The server estimates the probability of the disease

[1929] Input: Filtered case information.

[1930] How it works: The server uses statistical analysis and machine learning algorithms to estimate disease probabilities based on filtered case information, such as "the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%."

[1931] Output: Estimated probability information.

[1932] Step 8:

[1933] The server sends the estimation results to the device.

[1934] Input: Estimated probability information.

[1935] How it works: The server converts the probability information into JSON format and sends it to the device. The transmission process is secure using the SSL / TLS protocol.

[1936] Output: Probability information sent to the terminal.

[1937] Step 9:

[1938] The device displays the estimated results.

[1939] Input: Probability information sent to the terminal.

[1940] Operation: The device visually displays the received probability information in a list or graph format designed to be easily understood by the user.

[1941] Output: Probability information displayed to the user.

[1942] Step 10:

[1943] The server automatically generates the document

[1944] Input: Estimated probability information and basic patient information.

[1945] How it works: The server prepares templates for medical certificates and prescriptions based on the inference results, embeds the patient's individual information, and automatically generates draft documents using a template engine.

[1946] Output: Auto-generated draft document.

[1947] Step 11:

[1948] The device edits and displays the document.

[1949] Input: Auto-generated draft document.

[1950] Operation: The terminal displays the generated draft document to the user and provides an interface that allows the user to make corrections and additions. Once editing is complete, the user performs a final check and finalizes the document.

[1951] Output: Final revised and verified document.

[1952] Step 12:

[1953] The server stores the data and updates the learning model.

[1954] Input: Finalized documentation and diagnostic results.

[1955] How it works: The server stores the final diagnosis and treatment progress in a database. The stored data is used to periodically retrain the machine learning model to improve the accuracy of the system. The machine learning algorithm includes adaptive learning to accommodate increasing data.

[1956] Output: Updated machine learning model and saved diagnostic data.

[1957] (Application example 1)

[1958] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1959] In modern factories, maintenance and inspection of each machine is still often done manually. This relies on the knowledge and experience of the maintenance personnel, which can delay early detection of abnormalities and optimal maintenance responses. Furthermore, updating maintenance histories and manuals is time-consuming, which tends to lower the efficiency of the entire factory. Furthermore, identifying the cause of an abnormality and selecting a countermeasure can take time, so an efficient system to solve these problems is needed.

[1960] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1961] In this invention, the server includes input means for inputting patient information, means for generating keywords based on the input patient information, search means for retrieving case information from an external database using the generated keywords, analysis means for estimating the probability of disease based on the retrieved case information, means for displaying the estimated probability information, input means for inputting each device information, means for generating keywords based on the input device information, search means for retrieving maintenance response actions from an external database using the generated keywords, and means for applying the retrieved maintenance response actions to machine maintenance and repair. This makes it possible to efficiently and quickly identify machine abnormalities and propose optimal maintenance responses.

[1962] "Patient information" refers to basic patient data (such as name, age, sex, past medical history, specific symptoms, and test results) entered by medical professionals.

[1963] "Input means" refers to equipment or devices (PCs, tablets, smartphones, etc.) used to collect patient information and information on machinery and equipment and input it as data.

[1964] "Keyword generation means" refers to the technology or algorithm used to extract key keywords from input information.

[1965] A "search means" is a system or software that uses the generated keywords to retrieve related information from an external database.

[1966] "Analysis means" refers to a device or program that has the function of estimating and displaying specific results using statistical analysis or machine learning based on acquired information.

[1967] "Display means" refers to a display or monitor for visually displaying analysis results, estimated probability information, maintenance response actions, etc.

[1968] "Each device information" refers to data on the status and abnormalities of various machinery and equipment within the factory (such as years of operation, error logs, and current operating status).

[1969] "Maintenance response actions" refer to specific procedures and tasks for resolving abnormalities or breakdowns in machinery and equipment.

[1970] "Document generation means" refers to software or a system for automatically generating documents such as medical certificates, prescriptions, and maintenance manuals based on acquired information.

[1971] "Learning" is the process of using stored data to train machine learning models to improve the system's analytical accuracy and predictive capabilities.

[1972] This invention relates to a "smart maintenance system" for improving the efficiency of maintenance and inspection work for factory robots and machinery equipment. In particular, it is a system for automating a series of processes including input of equipment information, keyword generation, related information search, analysis, display, document generation, and learning.

[1973] 1. Entering device information and initial processing

[1974] User

[1975] Users (maintenance personnel) use dedicated devices (smartphones, tablets, etc.) to input information about each piece of machinery and equipment. For example, they input data such as the machine's name, years of operation, current status, and error logs. This information plays an important role in subsequent analysis and maintenance measures.

[1976] Terminal

[1977] The terminal converts the input data into a standard format and transmits it to the server using secure encrypted communications. After transmitting the data, the terminal notifies the user that the transmission has been completed.

[1978] 2. Data Acquisition and Analysis

[1979] server

[1980] The server stores the machine equipment data sent from the terminal in a database. At the same time, it checks the data for consistency. Next, it uses natural language processing (NLP) to generate keywords from the input information. For example, it extracts key keywords from error logs, such as "loud operating noise" or "vibration felt."

[1981] Based on the generated keywords, the server searches an external database for relevant maintenance actions, filtering the results based on criteria such as reliable data sources and keyword match.

[1982] 3. Estimation and provision of information on maintenance response actions

[1983] server

[1984] The server analyzes the filtered information and predicts the most appropriate maintenance action. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." The predicted maintenance action is sent to the terminal in JSON format.

[1985] Terminal

[1986] The terminal visually displays the received maintenance response actions. Possible display formats include lists, graphs, and text. In addition, documents such as maintenance manuals and maintenance histories are automatically generated based on the estimation results. The generated documents are displayed in a format that can be checked and edited by the user.

[1987] 4. Document Creation and Editing

[1988] server

[1989] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This document generation is performed based on predefined templates.

[1990] Terminal

[1991] The terminal displays the generated draft document to the user and provides an editing interface that allows the user to make corrections or additions as necessary. After the user checks and corrects the document, it is finalized.

[1992] 5. Data accumulation and accuracy improvement

[1993] server

[1994] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved.

[1995] Specific examples

[1996] The maintenance technician enters the status information of the "Robot Arm A1" (2 years of operation, operating noise, vibration detection) into the terminal. The input data is converted to JSON format and sent to the maintenance server.

[1997] The server analyzes the error log information sent and extracts key keywords ("operating sound" and "vibration") using TF-IDF. Maintenance response actions are searched and filtered based on the obtained keywords.

[1998] Based on the extracted keywords, the most appropriate maintenance actions ("Check the cooling system and clean the filters" and "Check for loose parts and lubricate the bearings") are displayed.

[1999] Prompt Sentence Examples

[2000] "You input the machine's condition, for example, 'it's loud when it's running' or 'I feel vibrations.' Then, based on this information, we suggest the best maintenance action."

[2001] As described above, the present invention provides a system that improves the efficiency of maintenance management within a factory, reduces the burden on maintenance personnel, and enables early detection of machine abnormalities.

[2002] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2003] Step 1:

[2004] Users input information about each piece of machinery and equipment using a dedicated device (smartphone, tablet, etc.). The input includes the machine's name, years of operation, current status, error log, etc. This allows data input (machine information) from the user to be obtained. This data is the basic data required for subsequent analysis.

[2005] Step 2:

[2006] The terminal converts the machine information entered by the user into a standard format, for example, JSON format, and sends it to the server using encrypted communication. This allows data to be sent from the terminal to the server.

[2007] Step 3:

[2008] The server receives the machine information sent from the terminal and stores it in a database. Before storing it, it checks the data for consistency and ensures that there are no abnormalities. This ensures the reliability of the data.

[2009] Step 4:

[2010] The server performs natural language processing (NLP) based on the received machine information to extract key keywords. For example, keywords such as "loud operating noise" and "vibration felt" are generated from error logs using TF-IDF. This provides basic keywords for analysis.

[2011] Step 5:

[2012] The server queries an external database (such as a maintenance manual database) based on the generated keywords. It extracts appropriate maintenance response actions and collects information on them. Search results are filtered based on reliable data sources, keyword matching, etc.

[2013] Step 6:

[2014] The server performs statistical analysis on the acquired maintenance response actions and estimates specific maintenance measures. For example, it recommends specific actions such as "inspect the cooling system and clean the filters" or "inspect loose parts and lubricate the bearings." This allows it to estimate the optimal maintenance response action.

[2015] Step 7:

[2016] The server converts the estimated maintenance action into JSON format and sends it to the terminal, which then transmits the analysis results to the terminal.

[2017] Step 8:

[2018] The terminal visually displays the received maintenance action and notifies the user. The display format can be a list, graph, text, etc. This allows the user to confirm the maintenance action.

[2019] Step 9:

[2020] The server prepares templates for maintenance manuals and maintenance histories based on the estimated maintenance actions, and automatically generates draft documents into which individual information for each device is inserted. This generates documents for maintenance actions.

[2021] Step 10:

[2022] The terminal displays the generated draft document to the user and provides an editing interface that allows corrections and additions as necessary. After the user checks and corrects the document, they finalize it. This finalizes the maintenance document.

[2023] Step 11:

[2024] The server stores the final maintenance results and maintenance history in a database. The stored data is then used to train a machine learning model. By periodically retraining the machine learning model with new data sets, the system's analytical accuracy and predictive capabilities are improved, thereby improving overall system performance.

[2025] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[2026] The present invention relates to an information provision system that efficiently inputs patient information and, based on that information, estimates disease probability, creates documents, and recognizes user emotions. The system includes an input means for inputting patient information, a keyword generation means, a search means, an analysis means, a display means, a document generation means, and a learning means. Furthermore, it incorporates an emotion engine and has the function of customizing the method of providing diagnosis results and advice based on the user's emotional state.

[2027] 1. Entering patient information and initial processing

[2028] User

[2029] The user (healthcare worker) logs in to a dedicated terminal and enters the patient's basic information (name, age, gender, etc.), specific symptoms (e.g., headache, fever, cough), past medical history, and current test results. Test results can be captured using a scanner or entered directly into the terminal.

[2030] Terminal

[2031] The terminal converts the input information into a standard format and transmits it to the server using encrypted communications. Once transmission is complete, the terminal notifies the user.

[2032] 2. Data Acquisition and Analysis

[2033] server

[2034] The server stores the patient information sent from the terminal in a database, checks the data consistency, and sends a supplemental request to the terminal if any information is missing.

[2035] server

[2036] The server analyzes stored patient information and generates keywords for related symptoms, using natural language processing (NLP) to extract key keywords from symptoms and test results.

[2037] server

[2038] The server searches online literature databases and case report sites based on the generated keywords, and retrieves relevant case information using medical database APIs and search engine queries.

[2039] server

[2040] The server performs filtering to extract highly relevant information from the acquired case information. Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[2041] server

[2042] The server performs statistical analysis based on the filtered case information to estimate the probability of the candidate disease. For example, it may estimate "40% probability of influenza," "30% probability of cold," or "20% probability of pneumonia." These estimation results are converted into JSON format and sent to the device.

[2043] 3. Information provision and document generation

[2044] Terminal

[2045] The device visually displays the received information so that the user can check it. It can use display formats such as lists and graphs. Furthermore, it automatically generates medical certificates, prescriptions, referral letters, etc. based on the inference results.

[2046] server

[2047] Based on the estimation results, the server prepares templates for medical certificates and prescriptions, inserts the patient's individual information, and generates draft documents.

[2048] Terminal

[2049] The terminal provides the generated draft document in an editable format, allowing the user to check and modify the draft document and make any necessary additions.

[2050] 4. Emotion Recognition and Response

[2051] Terminal

[2052] When the user operates the terminal, the terminal uses an emotion engine to recognize the user's emotional state from the user's facial expression, tone of voice, input content, etc.

[2053] server

[2054] The server evaluates the user's emotional state (e.g., stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[2055] server

[2056] Based on the assessment, the server adjusts the presentation of the diagnostic results to provide information in a format that is most acceptable to the user, for example, using calmer, more reassuring language for a user in a stressed state.

[2057] Terminal

[2058] The device uses an emotion engine to provide users with customized medical advice, for example, providing detailed instructions on specific steps to take or the need for additional testing if the user is worried.

[2059] 5. Data accumulation and learning

[2060] server

[2061] The server stores the final diagnosis and treatment progress in a database, which is then used to train a machine learning model to improve the accuracy of the system.

[2062] As a concrete example, if a patient complains of "headache, fever, and cough," the user enters this information into the terminal, and the server generates keywords and searches for cases based on "headache," "fever," and "cough." If the server estimates that the probability of influenza is 40%, the probability of a cold is 30%, and the probability of pneumonia is 20%, the terminal will display this information and further analyze the user's emotional state to provide information in the most optimal form. Medical certificates and prescriptions are also automatically generated, with the user making the final confirmation.

[2063] This system can provide information quickly and accurately, reducing the burden on medical professionals. Furthermore, emotion recognition can be used to present information in a way that is easy for users to accept, improving the quality of medical care.

[2064] The processing flow will be explained below.

[2065] Step 1: Enter patient information

[2066] User

[2067] The user (medical worker) logs in to a dedicated terminal.

[2068] The user enters the patient's basic information (name, age, gender, etc.), symptoms (headache, fever, cough, etc.), past medical history, and current test results.

[2069] Step 2: Sending data

[2070] Terminal

[2071] The terminal converts the entered patient information into a standard format.

[2072] The terminal transmits the patient information to the server using encrypted communication.

[2073] The terminal notifies the user that the transmission is complete.

[2074] Step 3: Save data and check its integrity

[2075] server

[2076] The server stores the received patient information in a database.

[2077] The server checks the integrity of the data and sends a supplemental request to the terminal if any information is missing.

[2078] Step 4: Generate keywords

[2079] server

[2080] The server analyzes stored patient information and automatically generates keywords for relevant symptoms.

[2081] Natural language processing (NLP) is used to extract key keywords from symptoms and test results.

[2082] Step 5: Internet Search

[2083] server

[2084] The server searches literature databases and case report sites on the Internet based on the generated keywords.

[2085] Use medical database APIs or search engine queries to retrieve relevant case information.

[2086] Step 6: Filtering case information

[2087] server

[2088] The server performs filtering to extract highly relevant information from the acquired case information.

[2089] Filtering criteria include the publication date of the case, reliable data source, and keyword matching.

[2090] Step 7: Estimate the probability of disease

[2091] server

[2092] The server performs statistical analysis based on the filtered case information and estimates the probability of each disease.

[2093] For example, estimates might say, "The probability of influenza is 40%," "The probability of a cold is 30%," and "The probability of pneumonia is 20%."

[2094] Step 8: Submit your information

[2095] server

[2096] The server converts the estimated probability information into JSON format and sends it to the terminal.

[2097] Step 9: Viewing Information

[2098] Terminal

[2099] The device visually displays the information it receives, for example in the form of a list or graph.

[2100] The user reviews and evaluates the displayed information.

[2101] Step 10: Emotion Recognition

[2102] Terminal

[2103] The device uses an emotion engine to analyze the user's facial expressions, tone of voice, and input content while operating the device, and recognizes the user's emotional state.

[2104] Step 11: Assess and regulate your emotional state

[2105] server

[2106] The server evaluates the user's emotional state (stress, relief, excitement, etc.) based on the analysis results of the emotion engine.

[2107] The server adjusts the way the information is presented based on the evaluation results, and provides the information in a format that is easy for the user to accept.

[2108] Step 12: Customize your information offering

[2109] Terminal

[2110] The terminal uses the emotion engine to present customized medical advice to the user.

[2111] For example, a user in a good emotional state may be provided with concise and efficient information, while a user experiencing stress may be provided with detailed and reassuring information.

[2112] Step 13: Document Generation

[2113] server

[2114] The server prepares templates for a diagnosis, prescription, and referral letter based on the estimation results and the user's emotional state.

[2115] Insert the patient's individual information into the template and automatically generate a draft document.

[2116] Step 14: Edit and review the document

[2117] Terminal

[2118] The terminal displays the generated draft document to the user and provides it in an editable format.

[2119] The user checks and corrects the draft document and makes any necessary additions.

[2120] Step 15: Save your data

[2121] server

[2122] The server stores the final diagnosis results and treatment progress in a database.

[2123] Step 16: Training the system and improving accuracy

[2124] server

[2125] The server uses the stored data to train the machine learning model.

[2126] The model is periodically retrained with new datasets to improve the accuracy of the system.

[2127] Example 2

[2128] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2129] In modern medical settings, entering and analyzing patient information, estimating diagnostic probabilities, and determining optimal treatment plans requires a great deal of time and effort. Furthermore, there is often a lack of information provided that takes into account the patient's emotions and mental state, which increases the burden on medical professionals and increases patient anxiety. To efficiently resolve these issues and improve the quality of medical care, a comprehensive and flexible information provision system is required.

[2130] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[2131] In this invention, the server includes means for inputting patient information, means for converting the input patient information into a standard format and transmitting it, means for saving the transmitted patient information, means for generating keywords based on the saved patient information, means for retrieving case information from an external database using the generated keywords, means for filtering the retrieved case information based on relevance, means for estimating disease probability based on the filtered case information, means for displaying the estimated probability information, and means for analyzing user emotions and adjusting the information presentation method, thereby enabling efficient input and analysis of patient information, accurate disease probability estimation, and information provision that takes patient emotions into consideration.

[2132] "Patient information" is a general term for basic patient information (such as name, age, and gender), specific symptoms, past medical history, and current test results.

[2133] "Means for inputting" refers to a device or interface that allows a user to input patient information.

[2134] "Means for converting into a standard format and transmitting" refers to the function for converting input patient information into a specified format and transmitting it securely to a server using an encrypted communication protocol.

[2135] "Means for storing" refers to a function for storing transmitted patient information in a database.

[2136] "Means for generating keywords" refers to technology that analyzes stored patient information and extracts keywords based on key symptoms and test results.

[2137] "Means for obtaining case information" refers to a function for obtaining related case information from external databases or case report sites using the generated keywords.

[2138] "Filtering means" refers to a function for sorting acquired case information based on relevance and extracting only highly reliable information.

[2139] "Means for estimating the probability of disease" refers to technology for performing statistical analysis based on filtered case information to calculate the probability of occurrence of a specific disease.

[2140] "Means for displaying probability information" refers to a display interface that allows a user to visually confirm the estimated disease occurrence probability.

[2141] "Means for analyzing emotions and adjusting the way information is presented" refers to a function that analyzes the user's emotional state, such as facial expressions and tone o...

Claims

1. an input means for inputting patient information; A means for generating keywords based on input patient information; a search means for acquiring case information from an external database using the generated keywords; an analysis means for estimating the probability of a disease based on the acquired case information; means for displaying the estimated probability information; The system that includes this.

2. a document generation means for automatically generating documents such as a medical certificate, a prescription, and a letter of introduction; means for displaying the generated document in an editable format; The system of claim 1 .

3. A learning method that uses the stored data to train a machine learning model to improve the accuracy of the system; and The system of claim 1 .

Images