System

Abstract

A system is provided.SOLUTION: The system includes a means for recording a conversation with a patient as voice data, a voice recognition means for converting the voice data into text data, a natural language processing means for extracting important information from the text data, a summary generation means for generating a summary based on the extracted important information, a means for providing the generated summary to a doctor or a person in charge, and a means for confirming and correcting the provided summary.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 modern medical institutions, patients often have to wait long periods of time to receive a diagnosis. One of the reasons for this is the time doctors spend writing up medical records after consultations. For this reason, it is necessary to shorten patient waiting times and improve the efficiency of medical treatment processes. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by providing a system that includes a means for recording conversations with patients as voice data, a speech recognition means for converting the voice data into text data, a natural language processing means for extracting important information from the text data, a summary generation means for generating a summary based on the extracted important information, a means for providing the generated summary to a doctor or other medical personnel, and a means for confirming and correcting the provided summary. This system generates medical information directly from the voice, reducing the burden on doctors and significantly shortening waiting times.

[0006] A "patient" is a person receiving medical care or treatment.

[0007] A conversation is an exchange of words between two or more people.

[0008] "Voice data" refers to data that is a digital recording of a human voice.

[0009] "Recording" means saving sound in a reproducible form.

[0010] "Speech recognition" is a technology that analyzes speech and converts it into text or instructions.

[0011] "Text data" is data in which character information is expressed in digital form.

[0012] "Natural language processing" is a technology that enables computers to understand and use human language.

[0013] "Extracting information" means extracting necessary information from a large amount of data.

[0014] A summary is a concise summary of a lot of information.

[0015] "Summary generation" refers to creating a concise summary based on the extracted information.

[0016] "Providing" means giving or allowing others to use information or services.

[0017] "Verification" means checking whether information or data is correct or not.

[0018] "Correction" means correcting an error or deficiency. [Brief explanation of the drawings]

[0019] [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

[0020] 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.

[0021] First, the terms used in the following description will be explained.

[0022] 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).

[0023] 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.

[0024] 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.

[0025] 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.

[0026] 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."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 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.

[0030] 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).

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 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.

[0037] 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.

[0038] 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.

[0039] 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."

[0040] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention will now be described.

[0041] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to the server in real time.

[0042] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[0043] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[0044] After extracting the key information, the server uses this information to generate a summary. This is done using a template-based summary generation algorithm. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the patient's main symptoms, when the symptoms began, and their medical history.

[0045] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[0046] Specific examples

[0047] For example, consider the case where a patient states:

[0048] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[0049] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." The text data is analyzed using a natural language processing engine, and information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold" is extracted. The server then uses this information to generate a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[0050] In this way, the present invention enables efficient and accurate generation of medical records, reduces the burden on doctors, and significantly reduces patient waiting times.

[0051] The processing flow will be explained below.

[0052] Step 1: Start recording

[0053] User: Patient and doctor begin a consultation and conversation.

[0054] The device uses a microphone to record this conversation as audio data.

[0055] The recorded audio data is temporarily stored in a local buffer.

[0056] Step 2: Sending audio data

[0057] As the conversation progresses, the device streams audio data in real time to the server.

[0058] The streaming audio data is continuously received by the server.

[0059] Step 3: Speech recognition processing

[0060] The server passes the received voice data to a voice recognition API and converts it into text data.

[0061] The speech recognition API analyzes the audio data and generates a corresponding string.

[0062] The generated text data is temporarily stored for the next processing step.

[0063] Step 4: Analyzing the text data

[0064] The server uses a natural language processing engine to analyze the text data.

[0065] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[0066] The extracted information is organized as structured data.

[0067] Step 5: Generate a summary

[0068] The server generates a summary based on the structured data.

[0069] A template-based summary generation algorithm is used to create a summary by applying the extracted information to a template sentence.

[0070] The generated summary includes the patient's main symptoms and the duration of symptoms.

[0071] Step 6: Provide a summary

[0072] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0073] The terminal receives the summary data and displays a screen that the physician can review and modify.

[0074] Step 7: Check and correct

[0075] User: The physician reviews the summary and makes corrections as needed through the interface.

[0076] The modified summary is then sent back to the server from the terminal.

[0077] Step 8: Save

[0078] The server receives the revised summary and stores it as the final data in the patient's medical record.

[0079] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[0080] This will automatically generate the necessary medical information from the voice data, reducing the burden on doctors and making the medical treatment process more efficient.

[0081] Example 1

[0082] 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."

[0083] In modern medical settings, it is important for medical professionals to quickly and accurately record conversations with patients and efficiently create medical records. However, manual recording is time-consuming and labor-intensive, and there is a risk of errors. In addition, there is a lack of systems that can comprehensively convert voice data into text data and extract and summarize important information. Therefore, a solution to this problem is needed.

[0084] 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.

[0085] In this invention, the server includes a device that records conversations with patients as voice data, a voice recognition device that converts the voice data into text data, and a natural language processing device that extracts important information from the text data. This allows for the creation of quick and accurate medical records by including a device that streams the voice data over the Internet, a device that saves the extracted information in JSON format or as a database record, a device that uses a dynamic template engine to generate summaries, and a device that saves the summaries in an electronic medical record system. This reduces the burden on medical professionals and shortens patient waiting times.

[0086] A "device for recording conversations with patients as audio data" is a device for capturing and recording conversations between patients and doctors as audio data using a microphone or the like.

[0087] A "voice recognition device that converts voice data into text data" is software or hardware that analyzes recorded voice data and converts it into corresponding text data.

[0088] A "natural language processing device that extracts important information from text data" is a device that includes algorithms and models for analyzing text data and automatically extracting necessary information such as a patient's symptoms and the duration of onset.

[0089] The "summary generation device that generates a summary based on extracted important information" refers to software or hardware that organizes the extracted important information and creates a summary using a template or the like.

[0090] The "apparatus for providing the generated summary to the medical professional" refers to a device or interface for displaying or linking the generated summary so that the medical professional can review it.

[0091] A "device for reviewing and correcting provided summaries" is an interface or software that allows a healthcare professional to review the generated summaries and make corrections as necessary.

[0092] A "device for streaming audio data over the Internet" is a communication device for transmitting recorded audio data to a server in real time.

[0093] The "device that stores extracted information in JSON format or as database records" is a software component that stores extracted important information as structured data.

[0094] An "apparatus using a dynamic template engine for generating summaries" is a software tool for dynamically applying extracted data to templates to generate summaries.

[0095] The "device that links and saves summaries in an electronic medical record system" refers to an interface or software that saves the generated summaries in an electronic medical record system and links them with the necessary medical information.

[0096] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention are described in detail below.

[0097] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and processes it with a digital signal processor (DSP) to obtain high-quality audio data. The recorded audio data is temporarily stored in a local buffer and then streamed to a server via the Internet. A secure communication protocol such as HTTPS is used to securely transmit the audio data.

[0098] The server processes the received voice data. Specifically, it converts the voice data into text data using a voice recognition device. In this case, common cloud services such as Google® Cloud Speech-to-Text and Amazon Transcribe can be used as the voice recognition API. The converted text data is temporarily stored in the server's database.

[0099] Next, a natural language processing system on the server analyzes this text data. A common natural language processing framework, such as spaCy or NLTK, is used as the natural language processing engine. The text data is first tokenized, and a key phrase extraction algorithm specialized for medical document analysis is applied to extract important information (e.g., symptoms, duration of symptom occurrence, relevant medical history, etc.). The extracted information is stored on the server in JSON format or as a database record.

[0100] The server then uses a summary generator to generate a summary based on the extracted key information. A template-based summary generation algorithm is used. Specifically, a dynamic template engine (e.g., Jinja2) is used to match the extracted data with specific template sentences to create a concise and accurate summary. The generated summary sentences are stored in the server's database.

[0101] The generated summary is sent from the server to the doctor's terminal. The summary is displayed through a user interface, where the user (doctor) can confirm it. The user interface is interactive, allowing the doctor to easily make corrections to the summary. The corrected summary is sent back to the server and linked to the electronic medical record system, where it is saved as the final medical record.

[0102] Specific examples

[0103] For example, consider the case where a patient states:

[0104] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[0105] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text using a speech recognition API, resulting in the sentence, "For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold." The natural language processing engine analyzes this text data and extracts information such as "for a week," "every night," "cough," and "it gets worse when it's cold." The server then uses this information to generate a summary.

[0106] "I started coughing every night for the past week, and it gets worse when it's cold."

[0107] The doctor then checks it and saves it in the medical record.

[0108] Example prompts for generative AI models

[0109] Examples of prompts for using an AI model to generate summaries include:

[0110] "Summarize the following conversation and include the main symptom and its duration: 'For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold.'"

[0111] This system enables efficient and accurate creation of medical records by integrating the conversion of voice data into text data, extraction of important information, generation of summaries, provision and correction of the summaries to medical professionals, and storage of the summaries as medical records.

[0112] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0113] Step 1:

[0114] The terminal records conversations between patients and doctors in real time. The input is the audio of the conversation during the consultation, and the output is audio data temporarily stored in a local buffer. Specifically, the terminal uses a microphone to capture the audio and processes the audio data via a digital signal processor (DSP), resulting in clear audio data with little noise.

[0115] Step 2:

[0116] The device streams recorded audio data to the server. The input is audio data stored in a local buffer, and the output is audio data sent to the server in real time. Specifically, the device uses an Internet connection to stream data using a secure communication protocol such as HTTPS. This ensures that the audio data is securely sent to the server.

[0117] Step 3:

[0118] The server processes the received voice data using a speech recognition API and converts it into text data. The input is streamed voice data, and the output is converted text data. The server then calls a speech recognition API (such as Google Cloud Speech-to-Text or Amazon Transcribe) to convert the voice data into text. This allows the content of the conversation to be obtained as text information.

[0119] Step 4:

[0120] The server analyzes the text data using a natural language processing engine to extract important information. The input is the text data, and the output is the extracted information (e.g., symptoms, duration of onset, relevant medical history). Specifically, the server uses a natural language processing engine (e.g., spaCy, NLTK) to tokenize the text data and apply a key phrase extraction algorithm. This allows the necessary medical information to be accurately extracted from the text.

[0121] Step 5:

[0122] The server generates a summary based on the extracted information. The input is the extracted key information, and the output is the generated summary. The server uses a template-based summary generation algorithm and a dynamic template engine (e.g., Jinja2) to match the extracted data with the defined template, resulting in the generated summary.

[0123] Step 6:

[0124] The server sends the generated summary to the doctor's terminal. The input is the generated summary text, and the output is the summary text displayed on the doctor's terminal. Specifically, the server sends the summary text to the doctor's terminal and displays it through the user interface, allowing the doctor to immediately check the summary content.

[0125] Step 7:

[0126] User: The doctor reviews the summary and makes corrections if necessary. The input is the displayed summary, and the output is the corrected summary. Specifically, the doctor makes corrections to the summary using an interactive user interface and sends the corrections to the server, which then produces the final confirmed summary.

[0127] Step 8:

[0128] The server connects the revised summary to the electronic medical record system and stores it. The input is the revised summary, and the output is the medical record stored in the electronic medical record system. The server integrates the revised summary into the electronic medical record system and stores it as a medical record, enabling long-term management of medical data.

[0129] This provides a system that can efficiently generate and manage useful and accurate clinical data from a series of sessions.

[0130] (Application example 1)

[0131] 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."

[0132] In factory production management, there is a need to efficiently collect and convert voice instructions from workers and managers into text, extract important information, and generate summaries. However, conventional methods require a great deal of time and effort to convert voice instructions into text and extract information, resulting in a significant decrease in work efficiency. In particular, in large factories, delays and misunderstandings in work instructions are likely to occur, adversely affecting quality control and productivity.

[0133] 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.

[0134] In this invention, the server includes means for recording conversations with workers or managers as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important instructions from the text data, summary generation means for generating a summary based on the extracted important instructions, means for providing the generated summary to the worker or manager, and means for checking and correcting the provided summary, thereby enabling voice instructions to be quickly and accurately converted into text and important information to be efficiently extracted and summarized.

[0135] "Workers" refer to employees who perform production and maintenance work within a factory.

[0136] A "manager" is a person in a position to manage business operations and give work instructions within a factory.

[0137] "Voice data" refers to data that has been digitally recorded from the voices of workers and managers.

[0138] "Speech recognition means" refers to a technique or device for converting voice data into text data.

[0139] "Text data" refers to data that has been converted into character information based on voice data.

[0140] "Natural language processing means" refers to technology or devices that extract meaningful information from text data.

[0141] A "summary generator" is a technique or device that generates a concise summary based on the extracted information.

[0142] A "summary" is a document that briefly summarizes important instructions or information.

[0143] The "presentation means" refers to a technique or device that presents the generated summary to the worker or manager.

[0144] "Verification and correction means" refers to techniques or devices that allow an operator or manager to verify the provided summary and make corrections as necessary.

[0145] The present invention is a system for efficiently collecting voice instructions from workers and managers in a production management system for factory robots, converting the voice data into text data, and extracting important instruction content and generating summaries. Specific embodiments for carrying out the present invention will be described below.

[0146] First, when a worker or manager gives instructions or reports by voice in the factory, the device records this conversation as voice data in real time. The hardware used can be a microphone, smart glasses, or a head-mounted display. The recorded voice data is temporarily stored in a local buffer and simultaneously streamed to the server.

[0147] The server processes the received voice data. First, the server converts the voice data into text data using a speech recognition API. Typical speech recognition APIs include Google Speech-to-Text and IBM Watson® Speech to Text. The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained generative AI model to extract important instructions (e.g., task details, deadlines, and important points) from the text data.

[0148] Specifically, the text data is tokenized and a key phrase extraction algorithm specialized for analyzing factory work instructions is applied. Natural language processing frameworks such as spaCy and NLTK are often used. After the important instruction content is extracted, the server generates a summary based on this information. A template-based summary generation algorithm is used to generate the summary. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the work content, deadlines, and important points.

[0149] The generated summary is sent from the server to the worker's or manager's terminal, where it can be viewed by the user through the terminal's user interface. The user, that is, the worker or manager, can review this summary and easily make corrections as necessary. An interface for this purpose is also provided. The corrected summary is finally saved in the production management system or work instruction system. This saving process is carried out in conjunction with the central management system.

[0150] As a specific example, consider the case where a worker states the following:

[0151] Worker: "Please complete the maintenance on Line 1 by tonight. Please be careful not to affect the operation of Line 2."

[0152] The terminal records this voice instruction and sends it as voice data to the server. The server converts the voice data into text, saying, "Please complete maintenance on Line 1 by tonight. Please take care not to affect the operation of Line 2." This text data is analyzed using a natural language processing engine, and information such as "tonight," "Line 1," "maintenance," and "operation of Line 2" is extracted. The server then uses this to generate a summary: "Complete maintenance on Line 1 by tonight. Take care to keep Line 2 running," and provides it to the user. The user then confirms this and saves it in the work instruction system.

[0153] Example prompts to input to a generative AI model:

[0154] Summarize the audio instructions below.

[0155] Voice command: "Please complete maintenance on Line 1 by this evening. Please be careful not to affect the operation of Line 2."

[0156] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0157] Step 1:

[0158] The terminal collects voice instructions from workers and managers using a microphone. The voice data is temporarily stored in a local buffer. The input is the voice of the worker or manager, and the output is voice data. The voice is captured using a microphone, converted into digital format, and stored.

[0159] Step 2:

[0160] The device streams locally stored audio data to the server in real time. The input is the audio data stored on the device, and the output is the audio data sent to the server. In this step, the audio data is sent to the server using an internet connection.

[0161] Step 3:

[0162] The server converts the received voice data into text data using a speech recognition API. The input is voice data and the output is text data. Specifically, APIs such as Google Speech-to-Text and IBM Watson Speech to Text are called to convert voice to text.

[0163] Step 4:

[0164] The server analyzes the text data using a natural language processing engine and extracts important instructions. The input is text data, and the output is important instructions (key phrases). Here, natural language processing frameworks such as spaCy and NLTK are used to tokenize the text and extract key phrases.

[0165] Step 5:

[0166] The server creates a summary based on the extracted key instructions using a template-based summary generation algorithm. The input is the key instructions, and the output is a summary. By embedding the extracted information in a template, a concise and accurate summary is generated.

[0167] Step 6:

[0168] The server sends the generated summary to the worker or administrator's terminal. The input is the summary text, and the output is the summary text displayed on the terminal. The sent summary text is delivered to the terminal via an Internet connection.

[0169] Step 7:

[0170] The terminal displays the summary to the worker or manager through a user interface. The input is the summary sent from the server, and the output is the displayed summary that the user can check. This step uses devices such as smart glasses or a head-mounted display.

[0171] Step 8:

[0172] The user, a worker or administrator, checks the provided summary and easily modifies it if necessary. The input is the displayed summary, and the output is the modified summary. The function to edit the summary is provided through the user interface.

[0173] Step 9:

[0174] The revised summary is finally saved in the production control system or work instruction system. The input is the revised summary, and the output is the saved summary. In this step, the summary is saved in cooperation with the database system.

[0175] 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.

[0176] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts the conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[0177] A patient and a doctor start talking in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to a server in real time.

[0178] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[0179] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[0180] Furthermore, the server uses an emotion engine that analyzes voice data to recognize the user's emotional state. This emotion engine analyzes the tone, rhythm, and speed of the voice and uses algorithms to detect the user's emotional state (e.g., relief, anxiety, anger, sadness, etc.). The emotion engine works in parallel with the speech-to-text conversion process using the speech recognition API.

[0181] The emotional state of the user recognized by the emotion engine is saved together with the analysis results of the natural language processing engine and reflected in the generation process of the summary generator, which generates a summary that includes information related to the user's emotional state in addition to the regular summary.

[0182] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[0183] Specific examples

[0184] For example, consider the case where a patient states:

[0185] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[0186] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. I seem anxious," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[0187] In this way, by combining the emotion engine, more detailed and accurate medical records can be generated that take into account the patient's emotional state, contributing to supporting doctors in their medical treatment.

[0188] The processing flow will be explained below.

[0189] Step 1: Start recording

[0190] User: Patient and doctor begin a consultation and conversation.

[0191] The device uses a microphone to record this conversation as audio data.

[0192] The recorded audio data is temporarily stored in a local buffer.

[0193] Step 2: Sending audio data

[0194] As the conversation progresses, the device streams audio data in real time to the server.

[0195] The streaming audio data is continuously received by the server.

[0196] Step 3: Speech recognition processing

[0197] The server passes the received voice data to a voice recognition API and converts it into text data.

[0198] The speech recognition API analyzes the audio data and generates a corresponding string.

[0199] The generated text data is temporarily stored for the next processing step.

[0200] Step 4: Emotion recognition processing

[0201] The server sends the voice data to the emotion engine to analyze the user's emotional state.

[0202] The emotion engine analyzes the tone, rhythm, and speed of speech to identify the user's emotional state (e.g., relief, anxiety, anger, sadness).

[0203] The emotional state recognized by the emotion engine is stored together with the text data.

[0204] Step 5: Analyzing the text data

[0205] The server uses a natural language processing engine to analyze the text data.

[0206] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[0207] The extracted information is organized as structured data.

[0208] Step 6: Generate a summary

[0209] The server generates a summary based on the structured data and the emotional state.

[0210] A template-based summary generation algorithm is used to create a summary sentence by applying the extracted information and emotional state to a template sentence.

[0211] The generated summary includes the patient's main symptoms, duration of symptoms, and emotional state.

[0212] Step 7: Provide a summary

[0213] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0214] The terminal receives the summary data and displays a screen that the physician can review and modify.

[0215] Step 8: Check and correct

[0216] User: The physician reviews the summary and makes corrections as needed through the interface.

[0217] The modified summary is then sent back to the server from the terminal.

[0218] Step 9: Save

[0219] The server receives the revised summary and stores it as the final data in the patient's medical record.

[0220] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[0221] This not only automatically generates the necessary medical information from the voice data, but also provides medical information that takes into account the patient's emotional state, improving the quality of medical care.

[0222] Example 2

[0223] 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."

[0224] In modern medical settings, there is a need to efficiently collect conversational information with patients and generate accurate medical records. However, simply converting voice data into text, extracting important information, and generating summaries does not provide medical support that takes into account the patient's emotional state. This can lead to a lack of understanding of the patient's psychological state, which could result in a decline in the quality of medical care.

[0225] 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.

[0226] In this invention, the server includes an emotion recognition means for analyzing the emotional state of the patient, a means for reflecting the analysis result of the emotion recognition means in the summary generation means, and a means for providing the generated summary, thereby enabling the generation of detailed and accurate medical records that also take into account the emotional state of the patient.

[0227] "Patient" means a person receiving medical services.

[0228] "Audio data" refers to recorded audio stored in digital format.

[0229] "Text data" refers to data obtained by converting voice data into character information.

[0230] "Speech recognition means" refers to technology or a system that converts voice data into text data.

[0231] "Natural language processing means" refers to technologies and systems that analyze and extract important information from text data.

[0232] A "summary generation means" is a technology or system that creates a summary based on the extracted information.

[0233] "Emotion recognition means" refers to technology or a system that analyzes voice data or text data and recognizes the user's emotional state.

[0234] A "summary" is a piece of text or data that briefly summarizes important information or key points.

[0235] "Means for providing" refers to the technology or system that presents the generated summary to doctors or other medical personnel.

[0236] "Verification and correction means" refers to techniques or systems that review the provided summary and correct it if necessary.

[0237] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts that conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions.

[0238] System configuration

[0239] 1. Terminal

[0240] The device is a device with audio capture capabilities for recording conversations. The device is equipped with a high-quality microphone, which records conversations in real time and temporarily stores them in a local buffer. The recorded audio data is streamed from the device to a server.

[0241] Specifically, the device used can be a smartphone or a dedicated recording device, and the start of recording can be triggered by, for example, pressing a button or using a voice command.

[0242] 2. Server

[0243] The server is composed of software with multiple functions, including voice recognition, natural language processing, and emotion recognition.

[0244] Specifically, it is composed as follows:

[0245] Speech recognition API: Use Google Cloud Speech-to-Text API or similar to convert voice data into text data.

[0246] Natural language processing engines: Use frameworks such as spaCy and NLTK to extract important information from text data.

[0247] Emotion Engine: Uses IBM Watson Tone Analyzer to analyze voice and text data and recognize the user's emotional state.

[0248] The server first receives the voice data and converts it into text data via a voice recognition API. It then analyzes the text data using a natural language processing engine to extract important information. At the same time, it analyzes the voice data using an emotion engine to recognize the user's emotional state. It then generates a summary based on the extracted important information and emotional state.

[0249] 3. User (Doctor)

[0250] The user, a physician, uses a terminal with an interface to review the generated summary and, if necessary, modify it. This interface displays the summary in a visually easy-to-understand format and allows editing.

[0251] The doctor can review the summary on the device and make any necessary corrections, after which the summary is sent back to the server and ultimately stored in the electronic medical record (EMR) system.

[0252] Specific examples

[0253] For example, consider the following situation where a patient states:

[0254] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[0255] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, he's had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "He started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. He seems anxious," and provides it to the doctor. The doctor then verifies this and saves it in the medical record.

[0256] This system will generate more detailed and accurate medical records that take into account the patient's emotional state, greatly assisting doctors in their medical practice.

[0257] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0258] Step 1:

[0259] When a conversation between a patient and a doctor begins, the device records the audio in real time using a microphone and temporarily stores it in a local buffer.

[0260] Specific operation: The patient and doctor start a conversation in the examination room, and the microphone on the device captures the audio. The device's button operation or voice command is used to trigger the start of recording.

[0261] Input: Voice of conversation between patient and doctor

[0262] Output: Locally stored audio data

[0263] Step 2:

[0264] The device streams the stored audio data to the server in real time.

[0265] Specific operation: The terminal maintains a stable network connection, divides the data stored in the buffer into certain packets, and delivers them to the server.

[0266] Input: Locally stored audio data

[0267] Output: Audio data streamed to the server

[0268] Step 3:

[0269] The server converts the received voice data into text data using a voice recognition API.

[0270] Specific operation: After receiving the voice data, the server sends the data to a speech recognition API and stores the returned text data in memory. For example, we use the Google Cloud Speech-to-Text API.

[0271] Input: Audio data streamed to the server

[0272] Output: Text data returned from the speech recognition API

[0273] Step 4:

[0274] The server analyzes the acquired text data using a natural language processing engine and extracts important information.

[0275] What it does: The server tokenizes the text data and applies algorithms (e.g., spaCy, NLTK) to extract key phrases such as symptoms, duration, and relevant medical history.

[0276] Input: Text data returned from the speech recognition API

[0277] Output: Key information extracted by the natural language processing engine

[0278] Step 5:

[0279] The server simultaneously inputs the voice data into an emotion recognition engine to recognize the patient's emotional state.

[0280] Specific operation: The server analyzes the tone, rhythm, speed, etc. of the voice data and identifies the emotional state (e.g., anxiety, relief, anger, sadness, etc.) using an emotion recognition algorithm (e.g., IBM Watson Tone Analyzer).

[0281] Input: Audio data streamed to the server

[0282] Output: Patient emotional state data from the emotion recognition engine

[0283] Step 6:

[0284] The server integrates the analyzed text data with the emotional state and generates a summary.

[0285] Specific operation: The server uses the extracted information and emotional state data to construct a summary sentence and stores it in memory.

[0286] Input: Important information extracted by a natural language processing engine, and patient emotional state data from an emotion recognition engine

[0287] Output: Generated summary

[0288] Step 7:

[0289] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0290] Specific operation: The generated summary is sent to the terminal and the summary content is displayed on the interface, including a visual display and editing function for doctors to check the summary.

[0291] Input: Generated summary

[0292] Output: Summary displayed on the terminal

[0293] Step 8:

[0294] User: The physician reviews the provided summary and makes any necessary corrections.

[0295] Specific actions: Check the summary on the device, make corrections in edit mode, and press the confirm button to save the changes.

[0296] Input: Summary text displayed on the terminal

[0297] Output: revised summary

[0298] Step 9:

[0299] The server finally saves the verified and corrected summary as a medical record in the electronic medical record (EMR) system.

[0300] Specific operation: The server calls the API of the EMR system and stores the revised summary data in association with the patient information.

[0301] Input: revised summary

[0302] Output: Medical records stored in the electronic medical record

[0303] (Application example 2)

[0304] 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."

[0305] A problem with modern smartphone usage is the lack of real-time security assessment based on the user's emotional state and conversation content. This makes it difficult to properly predict potential cyber threats and security risks and take countermeasures. Furthermore, the lack of technology to generate security alerts that reflect the user's unstable emotional state prevents users from using digital devices with peace of mind. To solve this problem, a fast and accurate security alert generation system that utilizes the user's conversation information and emotional state is needed.

[0306] 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.

[0307] In this invention, the server includes means for recording conversations with patients as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important information from the text data, summary generation means for generating a summary based on the extracted important information, means for providing the generated summary to a person in charge, means for confirming and correcting the provided summary, emotion analysis means for analyzing the emotional state, and means for creating and notifying a security alert based on the analysis results, thereby enabling real-time security evaluation and alert notification based on the user's conversation information and emotional state.

[0308] "Patient" refers to an individual receiving medical care or treatment.

[0309] "Audio data" refers to a data file that digitally records audio.

[0310] "Speech recognition means" refers to technology or devices for analyzing voice data and converting it into text data.

[0311] "Text data" refers to information expressed in text, typically stored in digital form.

[0312] "Natural language processing means" refers to technologies and devices for analyzing text data, understanding its meaning, and extracting important information.

[0313] "Significant information" refers to data that is particularly useful in relation to a particular purpose or use.

[0314] "Summary generator" refers to a technique or device that summarizes the extracted important information in a concise, easy-to-understand format.

[0315] "Contact Person" refers to the individual or group responsible for handling the generated summary and / or security assessment results.

[0316] "Emotion analysis means" refers to technology or devices for analyzing and determining a user's emotional state from voice or text data.

[0317] "Security Alert" refers to notifications intended to warn users of potential cyber threats or security risks.

[0318] "Means for creating and notifying security alerts" refers to technologies and devices for issuing security warnings to users based on the results of sentiment analysis and natural language processing means.

[0319] This invention is a system that analyzes a user's conversation information and emotional state, performs security assessments in real time, and generates and notifies appropriate security alerts.

[0320] First, the smartphone device records the conversation with the user as audio data. The audio data collected using the microphone is streamed to a server in real time. This communication uses the smartphone's internet connection. The collection and communication of this audio data uses general technology that is independent of the smartphone's operating system.

[0321] The server analyzes the received voice data. First, it uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) as a speech recognition method to convert the voice data into text data. This converted text data is then analyzed by a natural language processing method (e.g., spaCy). The natural language processing engine uses a pre-trained model to extract important information from the text data. Important information includes the user's activity, the period during which the activity occurred, and related historical information.

[0322] Next, the server uses an emotion analysis means to analyze the user's emotional state from the voice data. This emotion analysis takes into account the tone, rhythm, and speed of the voice. For example, an emotion analysis model from OpenAI (registered trademark) is used for the emotion analysis. Emotional states such as relief, anxiety, anger, and sadness are determined.

[0323] The analyzed text data and emotional state are combined into a single summary using a summary generator. The summary is generated using a template-based summary generation algorithm. The generated summary is provided to the human resource, who can review the summary and make any necessary revisions.

[0324] Furthermore, the server creates and notifies the user of security alerts based on the analysis results. The security alert evaluates potential cyber threats based on the user's emotional state and conversation content, and provides a warning accordingly. For example, if the emotional state is anxiety and the conversation content is "Could it be a virus?", the system will immediately notify the user with a recommendation to scan for a virus.

[0325] As a concrete example, consider the following user conversation:

[0326] "My phone has been running slow lately. Could this be due to a virus?" (Tone sounds concerned)

[0327] The smartphone records this conversation and sends it to the server. The server converts the voice into text, saying, "My smartphone has been running slowly lately. Could this be because I have a virus?" The text data is analyzed using a natural language processing engine, and information such as "My smartphone is running slowly" and "Possibly a virus" is extracted. In parallel, an emotion analysis tool analyzes the voice data and recognizes the user's emotional state as "worried." Based on this information, the server generates a summary: "My smartphone is running slowly, and I'm worried that it might have a virus." A person in charge checks this summary, and then the user is notified with a security alert, such as "We recommend a virus scan."

[0328] An example prompt for a generative AI model might look like this:

[0329] Input user conversation data as text: My smartphone has been running slow lately. Could it be because of a virus?

[0330] Enter the corresponding emotional state: Worry

[0331] Integrated results: Generate virus scanning recommendations based on conversation concerns.

[0332] The recommendation it generates is: Virus scanning and establishing security measures are recommended.

[0333] As a result, the present invention can propose security measures in real time based on the user's conversation information and emotional state, and provide an environment in which the user can use the device with peace of mind.

[0334] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0335] Step 1:

[0336] The smartphone device records the user's conversation. The user's conversation is acquired as audio data as input. Specifically, the smartphone's microphone is used to record the conversation in digital form. The audio data is generated as output.

[0337] Step 2:

[0338] The smartphone device streams audio data to a server in real time. Recorded audio data is used as input. Specifically, the smartphone's communication module is used to send the data to the server via the Internet. The server receives the audio data as output.

[0339] Step 3:

[0340] The server uses a speech recognition tool to convert the voice data into text data. The voice data streamed to the server is used as input. Specifically, a speech recognition API (e.g., Google Cloud Speech-to-Text API) is used to analyze and convert the voice data. Text data is generated as output.

[0341] Step 4:

[0342] The server uses natural language processing means to analyze the text data and extract important information. The text data generated by speech recognition is used as input. Specifically, a natural language processing engine (e.g., spaCy) is used to analyze the text data and extract important information (such as user activity, time period, and historical information). The important information is extracted as output.

[0343] Step 5:

[0344] The server uses emotion analysis means to analyze the user's emotional state from the voice data. The voice data is used as input. Specifically, an algorithm based on an emotion analysis model (e.g., OpenAI's emotion analysis model) is used to analyze the emotional state (relief, anxiety, anger, sadness, etc.) from the tone, rhythm, and speed of the voice. The analyzed emotional state is output.

[0345] Step 6:

[0346] The server uses a summary generation means to generate a summary based on the extracted key information and the analyzed emotional state. The key information and the emotional state are used as input. Specifically, a template-based summary generation algorithm is used to generate the summary. The summary is generated as output.

[0347] Step 7:

[0348] The server provides the generated summary to the person in charge. The generated summary is used as input. Specifically, the server sends the summary to the person in charge's terminal. As output, the summary is ready for the person in charge to check.

[0349] Step 8:

[0350] The person in charge checks the summary and makes corrections if necessary. The provided summary is used as input. Specifically, the person in charge checks the provided summary on their terminal and edits it using the correction interface. The corrected summary is generated as output.

[0351] Step 9:

[0352] The server creates a security alert based on the analysis results and notifies the user. The summary and sentiment analysis results are used as input. Specifically, an alert message is generated based on the analysis results and sent to the user's smartphone. The user receives the security alert as output.

[0353] 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.

[0354] 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.

[0355] 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.

[0356] [Second embodiment]

[0357] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0358] 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.

[0359] 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).

[0360] 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.

[0361] 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.

[0362] 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).

[0363] 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.

[0364] 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.

[0365] 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.

[0366] 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.

[0367] In the smart glasses 214, the 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.

[0368] 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."

[0369] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention will now be described.

[0370] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to the server in real time.

[0371] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[0372] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[0373] After extracting the key information, the server uses this information to generate a summary. This is done using a template-based summary generation algorithm. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the patient's main symptoms, when the symptoms began, and their medical history.

[0374] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[0375] Specific examples

[0376] For example, consider the case where a patient states:

[0377] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[0378] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." The text data is analyzed using a natural language processing engine, and information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold" is extracted. The server then uses this information to generate a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[0379] In this way, the present invention enables efficient and accurate generation of medical records, reduces the burden on doctors, and significantly reduces patient waiting times.

[0380] The processing flow will be explained below.

[0381] Step 1: Start recording

[0382] User: Patient and doctor begin a consultation and conversation.

[0383] The device uses a microphone to record this conversation as audio data.

[0384] The recorded audio data is temporarily stored in a local buffer.

[0385] Step 2: Sending audio data

[0386] As the conversation progresses, the device streams audio data in real time to the server.

[0387] The streaming audio data is continuously received by the server.

[0388] Step 3: Speech recognition processing

[0389] The server passes the received voice data to a voice recognition API and converts it into text data.

[0390] The speech recognition API analyzes the audio data and generates a corresponding string.

[0391] The generated text data is temporarily stored for the next processing step.

[0392] Step 4: Analyzing the text data

[0393] The server uses a natural language processing engine to analyze the text data.

[0394] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[0395] The extracted information is organized as structured data.

[0396] Step 5: Generate a summary

[0397] The server generates a summary based on the structured data.

[0398] A template-based summary generation algorithm is used to create a summary by applying the extracted information to a template sentence.

[0399] The generated summary includes the patient's main symptoms and the duration of symptoms.

[0400] Step 6: Provide a summary

[0401] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0402] The terminal receives the summary data and displays a screen that the physician can review and modify.

[0403] Step 7: Check and correct

[0404] User: The physician reviews the summary and makes corrections as needed through the interface.

[0405] The modified summary is then sent back to the server from the terminal.

[0406] Step 8: Save

[0407] The server receives the revised summary and stores it as the final data in the patient's medical record.

[0408] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[0409] This will automatically generate the necessary medical information from the voice data, reducing the burden on doctors and making the medical treatment process more efficient.

[0410] Example 1

[0411] 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."

[0412] In modern medical settings, it is important for medical professionals to quickly and accurately record conversations with patients and efficiently create medical records. However, manual recording is time-consuming and labor-intensive, and there is a risk of errors. In addition, there is a lack of systems that can comprehensively convert voice data into text data and extract and summarize important information. Therefore, a solution to this problem is needed.

[0413] 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.

[0414] In this invention, the server includes a device that records conversations with patients as voice data, a voice recognition device that converts the voice data into text data, and a natural language processing device that extracts important information from the text data. This allows for the creation of quick and accurate medical records by including a device that streams the voice data over the Internet, a device that saves the extracted information in JSON format or as a database record, a device that uses a dynamic template engine to generate summaries, and a device that saves the summaries in an electronic medical record system. This reduces the burden on medical professionals and shortens patient waiting times.

[0415] A "device for recording conversations with patients as audio data" is a device for capturing and recording conversations between patients and doctors as audio data using a microphone or the like.

[0416] A "voice recognition device that converts voice data into text data" is software or hardware that analyzes recorded voice data and converts it into corresponding text data.

[0417] A "natural language processing device that extracts important information from text data" is a device that includes algorithms and models for analyzing text data and automatically extracting necessary information such as a patient's symptoms and the duration of onset.

[0418] The "summary generation device that generates a summary based on extracted important information" refers to software or hardware that organizes the extracted important information and creates a summary using a template or the like.

[0419] The "apparatus for providing the generated summary to the medical professional" refers to a device or interface for displaying or linking the generated summary so that the medical professional can review it.

[0420] A "device for reviewing and correcting provided summaries" is an interface or software that allows a healthcare professional to review the generated summaries and make corrections as necessary.

[0421] A "device for streaming audio data over the Internet" is a communication device for transmitting recorded audio data to a server in real time.

[0422] The "device that stores extracted information in JSON format or as database records" is a software component that stores extracted important information as structured data.

[0423] An "apparatus using a dynamic template engine for generating summaries" is a software tool for dynamically applying extracted data to templates to generate summaries.

[0424] The "device that links and saves summaries in an electronic medical record system" refers to an interface or software that saves the generated summaries in an electronic medical record system and links them with the necessary medical information.

[0425] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention are described in detail below.

[0426] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and processes it with a digital signal processor (DSP) to obtain high-quality audio data. The recorded audio data is temporarily stored in a local buffer and then streamed to a server via the Internet. A secure communication protocol such as HTTPS is used to securely transmit the audio data.

[0427] The server processes the received voice data. Specifically, it converts the voice data into text data using a voice recognition device. Common cloud services such as Google Cloud Speech-to-Text and Amazon Transcribe can be used as the voice recognition API. The converted text data is temporarily stored in the server's database.

[0428] Next, a natural language processing system on the server analyzes this text data. A common natural language processing framework, such as spaCy or NLTK, is used as the natural language processing engine. The text data is first tokenized, and a key phrase extraction algorithm specialized for medical document analysis is applied to extract important information (e.g., symptoms, duration of symptom occurrence, relevant medical history, etc.). The extracted information is stored on the server in JSON format or as a database record.

[0429] The server then uses a summary generator to generate a summary based on the extracted key information. A template-based summary generation algorithm is used. Specifically, a dynamic template engine (e.g., Jinja2) is used to match the extracted data with specific template sentences to create a concise and accurate summary. The generated summary sentences are stored in the server's database.

[0430] The generated summary is sent from the server to the doctor's terminal. The summary is displayed through a user interface, where the user (doctor) can confirm it. The user interface is interactive, allowing the doctor to easily make corrections to the summary. The corrected summary is sent back to the server and linked to the electronic medical record system, where it is saved as the final medical record.

[0431] Specific examples

[0432] For example, consider the case where a patient states:

[0433] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[0434] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text using a speech recognition API, resulting in the sentence, "For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold." The natural language processing engine analyzes this text data and extracts information such as "for a week," "every night," "cough," and "it gets worse when it's cold." The server then uses this information to generate a summary.

[0435] "I started coughing every night for the past week, and it gets worse when it's cold."

[0436] The doctor then checks it and saves it in the medical record.

[0437] Example prompts for generative AI models

[0438] Examples of prompts for using an AI model to generate summaries include:

[0439] "Summarize the following conversation and include the main symptom and its duration: 'For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold.'"

[0440] This system enables efficient and accurate creation of medical records by integrating the conversion of voice data into text data, extraction of important information, generation of summaries, provision and correction of the summaries to medical professionals, and storage of the summaries as medical records.

[0441] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0442] Step 1:

[0443] The terminal records conversations between patients and doctors in real time. The input is the audio of the conversation during the consultation, and the output is audio data temporarily stored in a local buffer. Specifically, the terminal uses a microphone to capture the audio and processes the audio data via a digital signal processor (DSP), resulting in clear audio data with little noise.

[0444] Step 2:

[0445] The device streams recorded audio data to the server. The input is audio data stored in a local buffer, and the output is audio data sent to the server in real time. Specifically, the device uses an Internet connection to stream data using a secure communication protocol such as HTTPS. This ensures that the audio data is securely sent to the server.

[0446] Step 3:

[0447] The server processes the received voice data using a speech recognition API and converts it into text data. The input is streamed voice data, and the output is converted text data. The server then calls a speech recognition API (such as Google Cloud Speech-to-Text or Amazon Transcribe) to convert the voice data into text. This allows the content of the conversation to be obtained as text information.

[0448] Step 4:

[0449] The server analyzes the text data using a natural language processing engine to extract important information. The input is the text data, and the output is the extracted information (e.g., symptoms, duration of onset, relevant medical history). Specifically, the server uses a natural language processing engine (e.g., spaCy, NLTK) to tokenize the text data and apply a key phrase extraction algorithm. This allows the necessary medical information to be accurately extracted from the text.

[0450] Step 5:

[0451] The server generates a summary based on the extracted information. The input is the extracted key information, and the output is the generated summary. The server uses a template-based summary generation algorithm and a dynamic template engine (e.g., Jinja2) to match the extracted data with the defined template, resulting in the generated summary.

[0452] Step 6:

[0453] The server sends the generated summary to the doctor's terminal. The input is the generated summary text, and the output is the summary text displayed on the doctor's terminal. Specifically, the server sends the summary text to the doctor's terminal and displays it through the user interface, allowing the doctor to immediately check the summary content.

[0454] Step 7:

[0455] User: The doctor reviews the summary and makes corrections if necessary. The input is the displayed summary, and the output is the corrected summary. Specifically, the doctor makes corrections to the summary using an interactive user interface and sends the corrections to the server, which then produces the final confirmed summary.

[0456] Step 8:

[0457] The server connects the revised summary to the electronic medical record system and stores it. The input is the revised summary, and the output is the medical record stored in the electronic medical record system. The server integrates the revised summary into the electronic medical record system and stores it as a medical record, enabling long-term management of medical data.

[0458] This provides a system that can efficiently generate and manage useful and accurate clinical data from a series of sessions.

[0459] (Application example 1)

[0460] 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."

[0461] In factory production management, there is a need to efficiently collect and convert voice instructions from workers and managers into text, extract important information, and generate summaries. However, conventional methods require a great deal of time and effort to convert voice instructions into text and extract information, resulting in a significant decrease in work efficiency. In particular, in large factories, delays and misunderstandings in work instructions are likely to occur, adversely affecting quality control and productivity.

[0462] 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.

[0463] In this invention, the server includes means for recording conversations with workers or managers as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important instructions from the text data, summary generation means for generating a summary based on the extracted important instructions, means for providing the generated summary to the worker or manager, and means for checking and correcting the provided summary, thereby enabling voice instructions to be quickly and accurately converted into text and important information to be efficiently extracted and summarized.

[0464] "Workers" refer to employees who perform production and maintenance work within a factory.

[0465] A "manager" is a person in a position to manage business operations and give work instructions within a factory.

[0466] "Voice data" refers to data that has been digitally recorded from the voices of workers and managers.

[0467] "Speech recognition means" refers to a technique or device for converting voice data into text data.

[0468] "Text data" refers to data that has been converted into character information based on voice data.

[0469] "Natural language processing means" refers to technology or devices that extract meaningful information from text data.

[0470] A "summary generator" is a technique or device that generates a concise summary based on the extracted information.

[0471] A "summary" is a document that briefly summarizes important instructions or information.

[0472] The "presentation means" refers to a technique or device that presents the generated summary to the worker or manager.

[0473] "Verification and correction means" refers to techniques or devices that allow an operator or manager to verify the provided summary and make corrections as necessary.

[0474] The present invention is a system for efficiently collecting voice instructions from workers and managers in a production management system for factory robots, converting the voice data into text data, and extracting important instruction content and generating summaries. Specific embodiments for carrying out the present invention will be described below.

[0475] First, when a worker or manager gives instructions or reports by voice in the factory, the device records this conversation as voice data in real time. The hardware used can be a microphone, smart glasses, or a head-mounted display. The recorded voice data is temporarily stored in a local buffer and simultaneously streamed to the server.

[0476] The server processes the received voice data. First, the server converts the voice data into text data using a speech recognition API. Typical speech recognition APIs include Google Speech-to-Text and IBM Watson Speech to Text. The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained generative AI model to extract important instructions (e.g., work content, deadlines, points to note, etc.) from the text data.

[0477] Specifically, the text data is tokenized and a key phrase extraction algorithm specialized for analyzing factory work instructions is applied. Natural language processing frameworks such as spaCy and NLTK are often used. After the important instruction content is extracted, the server generates a summary based on this information. A template-based summary generation algorithm is used to generate the summary. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the work content, deadlines, and important points.

[0478] The generated summary is sent from the server to the worker's or manager's terminal, where it can be viewed by the user through the terminal's user interface. The user, that is, the worker or manager, can review this summary and easily make corrections as necessary. An interface for this purpose is also provided. The corrected summary is finally saved in the production management system or work instruction system. This saving process is carried out in conjunction with the central management system.

[0479] As a specific example, consider the case where a worker states the following:

[0480] Worker: "Please complete the maintenance on Line 1 by tonight. Please be careful not to affect the operation of Line 2."

[0481] The terminal records this voice instruction and sends it as voice data to the server. The server converts the voice data into text, saying, "Please complete maintenance on Line 1 by tonight. Please take care not to affect the operation of Line 2." This text data is analyzed using a natural language processing engine, and information such as "tonight," "Line 1," "maintenance," and "operation of Line 2" is extracted. The server then uses this to generate a summary: "Complete maintenance on Line 1 by tonight. Take care to keep Line 2 running," and provides it to the user. The user then confirms this and saves it in the work instruction system.

[0482] Example prompts to input to a generative AI model:

[0483] Summarize the audio instructions below.

[0484] Voice command: "Please complete maintenance on Line 1 by this evening. Please be careful not to affect the operation of Line 2."

[0485] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0486] Step 1:

[0487] The terminal collects voice instructions from workers and managers using a microphone. The voice data is temporarily stored in a local buffer. The input is the voice of the worker or manager, and the output is voice data. The voice is captured using a microphone, converted into digital format, and stored.

[0488] Step 2:

[0489] The device streams locally stored audio data to the server in real time. The input is the audio data stored on the device, and the output is the audio data sent to the server. In this step, the audio data is sent to the server using an internet connection.

[0490] Step 3:

[0491] The server converts the received voice data into text data using a speech recognition API. The input is voice data and the output is text data. Specifically, APIs such as Google Speech-to-Text and IBM Watson Speech to Text are called to convert voice to text.

[0492] Step 4:

[0493] The server analyzes the text data using a natural language processing engine and extracts important instructions. The input is text data, and the output is important instructions (key phrases). Here, natural language processing frameworks such as spaCy and NLTK are used to tokenize the text and extract key phrases.

[0494] Step 5:

[0495] The server creates a summary based on the extracted key instructions using a template-based summary generation algorithm. The input is the key instructions, and the output is a summary. By embedding the extracted information in a template, a concise and accurate summary is generated.

[0496] Step 6:

[0497] The server sends the generated summary to the worker or administrator's terminal. The input is the summary text, and the output is the summary text displayed on the terminal. The sent summary text is delivered to the terminal via an Internet connection.

[0498] Step 7:

[0499] The terminal displays the summary to the worker or manager through a user interface. The input is the summary sent from the server, and the output is the displayed summary that the user can check. This step uses devices such as smart glasses or a head-mounted display.

[0500] Step 8:

[0501] The user, a worker or administrator, checks the provided summary and easily modifies it if necessary. The input is the displayed summary, and the output is the modified summary. The function to edit the summary is provided through the user interface.

[0502] Step 9:

[0503] The revised summary is finally saved in the production control system or work instruction system. The input is the revised summary, and the output is the saved summary. In this step, the summary is saved in cooperation with the database system.

[0504] 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.

[0505] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts the conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[0506] A patient and a doctor start talking in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to a server in real time.

[0507] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[0508] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[0509] Furthermore, the server uses an emotion engine that analyzes voice data to recognize the user's emotional state. This emotion engine analyzes the tone, rhythm, and speed of the voice and uses algorithms to detect the user's emotional state (e.g., relief, anxiety, anger, sadness, etc.). The emotion engine works in parallel with the speech-to-text conversion process using the speech recognition API.

[0510] The emotional state of the user recognized by the emotion engine is saved together with the analysis results of the natural language processing engine and reflected in the generation process of the summary generator, which generates a summary that includes information related to the user's emotional state in addition to the regular summary.

[0511] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[0512] Specific examples

[0513] For example, consider the case where a patient states:

[0514] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[0515] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. I seem anxious," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[0516] In this way, by combining the emotion engine, more detailed and accurate medical records can be generated that take into account the patient's emotional state, contributing to supporting doctors in their medical treatment.

[0517] The processing flow will be explained below.

[0518] Step 1: Start recording

[0519] User: Patient and doctor begin a consultation and conversation.

[0520] The device uses a microphone to record this conversation as audio data.

[0521] The recorded audio data is temporarily stored in a local buffer.

[0522] Step 2: Sending audio data

[0523] As the conversation progresses, the device streams audio data in real time to the server.

[0524] The streaming audio data is continuously received by the server.

[0525] Step 3: Speech recognition processing

[0526] The server passes the received voice data to a voice recognition API and converts it into text data.

[0527] The speech recognition API analyzes the audio data and generates a corresponding string.

[0528] The generated text data is temporarily stored for the next processing step.

[0529] Step 4: Emotion recognition processing

[0530] The server sends the voice data to the emotion engine to analyze the user's emotional state.

[0531] The emotion engine analyzes the tone, rhythm, and speed of speech to identify the user's emotional state (e.g., relief, anxiety, anger, sadness).

[0532] The emotional state recognized by the emotion engine is stored together with the text data.

[0533] Step 5: Analyzing the text data

[0534] The server uses a natural language processing engine to analyze the text data.

[0535] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[0536] The extracted information is organized as structured data.

[0537] Step 6: Generate a summary

[0538] The server generates a summary based on the structured data and the emotional state.

[0539] A template-based summary generation algorithm is used to create a summary sentence by applying the extracted information and emotional state to a template sentence.

[0540] The generated summary includes the patient's main symptoms, duration of symptoms, and emotional state.

[0541] Step 7: Provide a summary

[0542] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0543] The terminal receives the summary data and displays a screen that the physician can review and modify.

[0544] Step 8: Check and correct

[0545] User: The physician reviews the summary and makes corrections as needed through the interface.

[0546] The modified summary is then sent back to the server from the terminal.

[0547] Step 9: Save

[0548] The server receives the revised summary and stores it as the final data in the patient's medical record.

[0549] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[0550] This not only automatically generates the necessary medical information from the voice data, but also provides medical information that takes into account the patient's emotional state, improving the quality of medical care.

[0551] Example 2

[0552] 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."

[0553] In modern medical settings, there is a need to efficiently collect conversational information with patients and generate accurate medical records. However, simply converting voice data into text, extracting important information, and generating summaries does not provide medical support that takes into account the patient's emotional state. This can lead to a lack of understanding of the patient's psychological state, which could result in a decline in the quality of medical care.

[0554] 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.

[0555] In this invention, the server includes an emotion recognition means for analyzing the emotional state of the patient, a means for reflecting the analysis result of the emotion recognition means in the summary generation means, and a means for providing the generated summary, thereby enabling the generation of detailed and accurate medical records that also take into account the emotional state of the patient.

[0556] "Patient" means a person receiving medical services.

[0557] "Audio data" refers to recorded audio stored in digital format.

[0558] "Text data" refers to data obtained by converting voice data into character information.

[0559] "Speech recognition means" refers to technology or a system that converts voice data into text data.

[0560] "Natural language processing means" refers to technologies and systems that analyze and extract important information from text data.

[0561] A "summary generation means" is a technology or system that creates a summary based on the extracted information.

[0562] "Emotion recognition means" refers to technology or a system that analyzes voice data or text data and recognizes the user's emotional state.

[0563] A "summary" is a piece of text or data that briefly summarizes important information or key points.

[0564] "Means for providing" refers to the technology or system that presents the generated summary to doctors or other medical personnel.

[0565] "Verification and correction means" refers to techniques or systems that review the provided summary and correct it if necessary.

[0566] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts that conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions.

[0567] System configuration

[0568] 1. Terminal

[0569] The device is a device with audio capture capabilities for recording conversations. The device is equipped with a high-quality microphone, which records conversations in real time and temporarily stores them in a local buffer. The recorded audio data is streamed from the device to a server.

[0570] Specifically, the device used can be a smartphone or a dedicated recording device, and the start of recording can be triggered by, for example, pressing a button or using a voice command.

[0571] 2. Server

[0572] The server is composed of software with multiple functions, including voice recognition, natural language processing, and emotion recognition.

[0573] Specifically, it is composed as follows:

[0574] Speech recognition API: Use Google Cloud Speech-to-Text API or similar to convert voice data into text data.

[0575] Natural language processing engines: Use frameworks such as spaCy and NLTK to extract important information from text data.

[0576] Emotion Engine: Uses IBM Watson Tone Analyzer to analyze voice and text data and recognize the user's emotional state.

[0577] The server first receives the voice data and converts it into text data via a voice recognition API. It then analyzes the text data using a natural language processing engine to extract important information. At the same time, it analyzes the voice data using an emotion engine to recognize the user's emotional state. It then generates a summary based on the extracted important information and emotional state.

[0578] 3. User (Doctor)

[0579] The user, a physician, uses a terminal with an interface to review the generated summary and, if necessary, modify it. This interface displays the summary in a visually easy-to-understand format and allows editing.

[0580] The doctor can review the summary on the device and make any necessary corrections, after which the summary is sent back to the server and ultimately stored in the electronic medical record (EMR) system.

[0581] Specific examples

[0582] For example, consider the following situation where a patient states:

[0583] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[0584] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, he's had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "He started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. He seems anxious," and provides it to the doctor. The doctor then verifies this and saves it in the medical record.

[0585] This system will generate more detailed and accurate medical records that take into account the patient's emotional state, greatly assisting doctors in their medical practice.

[0586] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0587] Step 1:

[0588] When a conversation between a patient and a doctor begins, the device records the audio in real time using a microphone and temporarily stores it in a local buffer.

[0589] Specific operation: The patient and doctor start a conversation in the examination room, and the microphone on the device captures the audio. The device's button operation or voice command is used to trigger the start of recording.

[0590] Input: Voice of conversation between patient and doctor

[0591] Output: Locally stored audio data

[0592] Step 2:

[0593] The device streams the stored audio data to the server in real time.

[0594] Specific operation: The terminal maintains a stable network connection, divides the data stored in the buffer into certain packets, and delivers them to the server.

[0595] Input: Locally stored audio data

[0596] Output: Audio data streamed to the server

[0597] Step 3:

[0598] The server converts the received voice data into text data using a voice recognition API.

[0599] Specific operation: After receiving the voice data, the server sends the data to a speech recognition API and stores the returned text data in memory. For example, we use the Google Cloud Speech-to-Text API.

[0600] Input: Audio data streamed to the server

[0601] Output: Text data returned from the speech recognition API

[0602] Step 4:

[0603] The server analyzes the acquired text data using a natural language processing engine and extracts important information.

[0604] What it does: The server tokenizes the text data and applies algorithms (e.g., spaCy, NLTK) to extract key phrases such as symptoms, duration, and relevant medical history.

[0605] Input: Text data returned from the speech recognition API

[0606] Output: Key information extracted by the natural language processing engine

[0607] Step 5:

[0608] The server simultaneously inputs the voice data into an emotion recognition engine to recognize the patient's emotional state.

[0609] Specific operation: The server analyzes the tone, rhythm, speed, etc. of the voice data and identifies the emotional state (e.g., anxiety, relief, anger, sadness, etc.) using an emotion recognition algorithm (e.g., IBM Watson Tone Analyzer).

[0610] Input: Audio data streamed to the server

[0611] Output: Patient emotional state data from the emotion recognition engine

[0612] Step 6:

[0613] The server integrates the analyzed text data with the emotional state and generates a summary.

[0614] Specific operation: The server uses the extracted information and emotional state data to construct a summary sentence and stores it in memory.

[0615] Input: Important information extracted by a natural language processing engine, and patient emotional state data from an emotion recognition engine

[0616] Output: Generated summary

[0617] Step 7:

[0618] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0619] Specific operation: The generated summary is sent to the terminal and the summary content is displayed on the interface, including a visual display and editing function for doctors to check the summary.

[0620] Input: Generated summary

[0621] Output: Summary displayed on the terminal

[0622] Step 8:

[0623] User: The physician reviews the provided summary and makes any necessary corrections.

[0624] Specific actions: Check the summary on the device, make corrections in edit mode, and press the confirm button to save the changes.

[0625] Input: Summary text displayed on the terminal

[0626] Output: revised summary

[0627] Step 9:

[0628] The server finally saves the verified and corrected summary as a medical record in the electronic medical record (EMR) system.

[0629] Specific operation: The server calls the API of the EMR system and stores the revised summary data in association with the patient information.

[0630] Input: revised summary

[0631] Output: Medical records stored in the electronic medical record

[0632] (Application example 2)

[0633] 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."

[0634] A problem with modern smartphone usage is the lack of real-time security assessment based on the user's emotional state and conversation content. This makes it difficult to properly predict potential cyber threats and security risks and take countermeasures. Furthermore, the lack of technology to generate security alerts that reflect the user's unstable emotional state prevents users from using digital devices with peace of mind. To solve this problem, a fast and accurate security alert generation system that utilizes the user's conversation information and emotional state is needed.

[0635] 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.

[0636] In this invention, the server includes means for recording conversations with patients as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important information from the text data, summary generation means for generating a summary based on the extracted important information, means for providing the generated summary to a person in charge, means for confirming and correcting the provided summary, emotion analysis means for analyzing the emotional state, and means for creating and notifying a security alert based on the analysis results, thereby enabling real-time security evaluation and alert notification based on the user's conversation information and emotional state.

[0637] "Patient" refers to an individual receiving medical care or treatment.

[0638] "Audio data" refers to a data file that digitally records audio.

[0639] "Speech recognition means" refers to technology or devices for analyzing voice data and converting it into text data.

[0640] "Text data" refers to information expressed in text, typically stored in digital form.

[0641] "Natural language processing means" refers to technologies and devices for analyzing text data, understanding its meaning, and extracting important information.

[0642] "Significant information" refers to data that is particularly useful in relation to a particular purpose or use.

[0643] "Summary generator" refers to a technique or device that summarizes the extracted important information in a concise, easy-to-understand format.

[0644] "Contact Person" refers to the individual or group responsible for handling the generated summary and / or security assessment results.

[0645] "Emotion analysis means" refers to technology or devices for analyzing and determining a user's emotional state from voice or text data.

[0646] "Security Alert" refers to notifications intended to warn users of potential cyber threats or security risks.

[0647] "Means for creating and notifying security alerts" refers to technologies and devices for issuing security warnings to users based on the results of sentiment analysis and natural language processing means.

[0648] This invention is a system that analyzes a user's conversation information and emotional state, performs security assessments in real time, and generates and notifies appropriate security alerts.

[0649] First, the smartphone device records the conversation with the user as audio data. The audio data collected using the microphone is streamed to a server in real time. This communication uses the smartphone's internet connection. The collection and communication of this audio data uses general technology that is independent of the smartphone's operating system.

[0650] The server analyzes the received voice data. First, it uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) as a speech recognition method to convert the voice data into text data. This converted text data is then analyzed by a natural language processing method (e.g., spaCy). The natural language processing engine uses a pre-trained model to extract important information from the text data. Important information includes the user's activity, the period during which the activity occurred, and related historical information.

[0651] Next, the server uses emotion analysis to analyze the user's emotional state from the voice data. This emotion analysis takes into account the tone, rhythm, and speed of the voice. For example, OpenAI's emotion analysis model is used for emotion analysis. Emotional states such as relief, anxiety, anger, and sadness are determined.

[0652] The analyzed text data and emotional state are combined into a single summary using a summary generator. The summary is generated using a template-based summary generation algorithm. The generated summary is provided to the human resource, who can review the summary and make any necessary revisions.

[0653] Furthermore, the server creates and notifies the user of security alerts based on the analysis results. The security alert evaluates potential cyber threats based on the user's emotional state and conversation content, and provides a warning accordingly. For example, if the emotional state is anxiety and the conversation content is "Could it be a virus?", the system will immediately notify the user with a recommendation to scan for a virus.

[0654] As a concrete example, consider the following user conversation:

[0655] "My phone has been running slow lately. Could this be due to a virus?" (Tone sounds concerned)

[0656] The smartphone records this conversation and sends it to the server. The server converts the voice into text, saying, "My smartphone has been running slowly lately. Could this be because I have a virus?" The text data is analyzed using a natural language processing engine, and information such as "My smartphone is running slowly" and "Possibly a virus" is extracted. In parallel, an emotion analysis tool analyzes the voice data and recognizes the user's emotional state as "worried." Based on this information, the server generates a summary: "My smartphone is running slowly, and I'm worried that it might have a virus." A person in charge checks this summary, and then the user is notified with a security alert, such as "We recommend a virus scan."

[0657] An example prompt for a generative AI model might look like this:

[0658] Input user conversation data as text: My smartphone has been running slow lately. Could it be because of a virus?

[0659] Enter the corresponding emotional state: Worry

[0660] Integrated results: Generate virus scanning recommendations based on conversation concerns.

[0661] The recommendation it generates is: Virus scanning and establishing security measures are recommended.

[0662] As a result, the present invention can propose security measures in real time based on the user's conversation information and emotional state, and provide an environment in which the user can use the device with peace of mind.

[0663] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0664] Step 1:

[0665] The smartphone device records the user's conversation. The user's conversation is acquired as audio data as input. Specifically, the smartphone's microphone is used to record the conversation in digital form. The audio data is generated as output.

[0666] Step 2:

[0667] The smartphone device streams audio data to a server in real time. Recorded audio data is used as input. Specifically, the smartphone's communication module is used to send the data to the server via the Internet. The server receives the audio data as output.

[0668] Step 3:

[0669] The server uses a speech recognition tool to convert the voice data into text data. The voice data streamed to the server is used as input. Specifically, a speech recognition API (e.g., Google Cloud Speech-to-Text API) is used to analyze and convert the voice data. Text data is generated as output.

[0670] Step 4:

[0671] The server uses natural language processing means to analyze the text data and extract important information. The text data generated by speech recognition is used as input. Specifically, a natural language processing engine (e.g., spaCy) is used to analyze the text data and extract important information (such as user activity, time period, and historical information). The important information is extracted as output.

[0672] Step 5:

[0673] The server uses emotion analysis means to analyze the user's emotional state from the voice data. The voice data is used as input. Specifically, an algorithm based on an emotion analysis model (e.g., OpenAI's emotion analysis model) is used to analyze the emotional state (relief, anxiety, anger, sadness, etc.) from the tone, rhythm, and speed of the voice. The analyzed emotional state is output.

[0674] Step 6:

[0675] The server uses a summary generation means to generate a summary based on the extracted key information and the analyzed emotional state. The key information and the emotional state are used as input. Specifically, a template-based summary generation algorithm is used to generate the summary. The summary is generated as output.

[0676] Step 7:

[0677] The server provides the generated summary to the person in charge. The generated summary is used as input. Specifically, the server sends the summary to the person in charge's terminal. As output, the summary is ready for the person in charge to check.

[0678] Step 8:

[0679] The person in charge checks the summary and makes corrections if necessary. The provided summary is used as input. Specifically, the person in charge checks the provided summary on their terminal and edits it using the correction interface. The corrected summary is generated as output.

[0680] Step 9:

[0681] The server creates a security alert based on the analysis results and notifies the user. The summary and sentiment analysis results are used as input. Specifically, an alert message is generated based on the analysis results and sent to the user's smartphone. The user receives the security alert as output.

[0682] 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.

[0683] 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.

[0684] 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.

[0685] [Third embodiment]

[0686] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0687] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0688] 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).

[0689] 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.

[0690] 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.

[0691] 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).

[0692] 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.

[0693] 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.

[0694] 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.

[0695] 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.

[0696] 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.

[0697] 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."

[0698] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention will now be described.

[0699] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to the server in real time.

[0700] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[0701] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[0702] After extracting the key information, the server uses this information to generate a summary. This is done using a template-based summary generation algorithm. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the patient's main symptoms, when the symptoms began, and their medical history.

[0703] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[0704] Specific examples

[0705] For example, consider the case where a patient states:

[0706] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[0707] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." The text data is analyzed using a natural language processing engine, and information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold" is extracted. The server then uses this information to generate a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[0708] In this way, the present invention enables efficient and accurate generation of medical records, reduces the burden on doctors, and significantly reduces patient waiting times.

[0709] The processing flow will be explained below.

[0710] Step 1: Start recording

[0711] User: Patient and doctor begin a consultation and conversation.

[0712] The device uses a microphone to record this conversation as audio data.

[0713] The recorded audio data is temporarily stored in a local buffer.

[0714] Step 2: Sending audio data

[0715] As the conversation progresses, the device streams audio data in real time to the server.

[0716] The streaming audio data is continuously received by the server.

[0717] Step 3: Speech recognition processing

[0718] The server passes the received voice data to a voice recognition API and converts it into text data.

[0719] The speech recognition API analyzes the audio data and generates a corresponding string.

[0720] The generated text data is temporarily stored for the next processing step.

[0721] Step 4: Analyzing the text data

[0722] The server uses a natural language processing engine to analyze the text data.

[0723] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[0724] The extracted information is organized as structured data.

[0725] Step 5: Generate a summary

[0726] The server generates a summary based on the structured data.

[0727] A template-based summary generation algorithm is used to create a summary by applying the extracted information to a template sentence.

[0728] The generated summary includes the patient's main symptoms and the duration of symptoms.

[0729] Step 6: Provide a summary

[0730] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0731] The terminal receives the summary data and displays a screen that the physician can review and modify.

[0732] Step 7: Check and correct

[0733] User: The physician reviews the summary and makes corrections as needed through the interface.

[0734] The modified summary is then sent back to the server from the terminal.

[0735] Step 8: Save

[0736] The server receives the revised summary and stores it as the final data in the patient's medical record.

[0737] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[0738] This will automatically generate the necessary medical information from the voice data, reducing the burden on doctors and making the medical treatment process more efficient.

[0739] Example 1

[0740] 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."

[0741] In modern medical settings, it is important for medical professionals to quickly and accurately record conversations with patients and efficiently create medical records. However, manual recording is time-consuming and labor-intensive, and there is a risk of errors. In addition, there is a lack of systems that can comprehensively convert voice data into text data and extract and summarize important information. Therefore, a solution to this problem is needed.

[0742] 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.

[0743] In this invention, the server includes a device that records conversations with patients as voice data, a voice recognition device that converts the voice data into text data, and a natural language processing device that extracts important information from the text data. This allows for the creation of quick and accurate medical records by including a device that streams the voice data over the Internet, a device that saves the extracted information in JSON format or as a database record, a device that uses a dynamic template engine to generate summaries, and a device that saves the summaries in an electronic medical record system. This reduces the burden on medical professionals and shortens patient waiting times.

[0744] A "device for recording conversations with patients as audio data" is a device for capturing and recording conversations between patients and doctors as audio data using a microphone or the like.

[0745] A "voice recognition device that converts voice data into text data" is software or hardware that analyzes recorded voice data and converts it into corresponding text data.

[0746] A "natural language processing device that extracts important information from text data" is a device that includes algorithms and models for analyzing text data and automatically extracting necessary information such as a patient's symptoms and the duration of onset.

[0747] The "summary generation device that generates a summary based on extracted important information" refers to software or hardware that organizes the extracted important information and creates a summary using a template or the like.

[0748] The "apparatus for providing the generated summary to the medical professional" refers to a device or interface for displaying or linking the generated summary so that the medical professional can review it.

[0749] A "device for reviewing and correcting provided summaries" is an interface or software that allows a healthcare professional to review the generated summaries and make corrections as necessary.

[0750] A "device for streaming audio data over the Internet" is a communication device for transmitting recorded audio data to a server in real time.

[0751] The "device that stores extracted information in JSON format or as database records" is a software component that stores extracted important information as structured data.

[0752] An "apparatus using a dynamic template engine for generating summaries" is a software tool for dynamically applying extracted data to templates to generate summaries.

[0753] The "device that links and saves summaries in an electronic medical record system" refers to an interface or software that saves the generated summaries in an electronic medical record system and links them with the necessary medical information.

[0754] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention are described in detail below.

[0755] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and processes it with a digital signal processor (DSP) to obtain high-quality audio data. The recorded audio data is temporarily stored in a local buffer and then streamed to a server via the Internet. A secure communication protocol such as HTTPS is used to securely transmit the audio data.

[0756] The server processes the received voice data. Specifically, it converts the voice data into text data using a voice recognition device. Common cloud services such as Google Cloud Speech-to-Text and Amazon Transcribe can be used as the voice recognition API. The converted text data is temporarily stored in the server's database.

[0757] Next, a natural language processing system on the server analyzes this text data. A common natural language processing framework, such as spaCy or NLTK, is used as the natural language processing engine. The text data is first tokenized, and a key phrase extraction algorithm specialized for medical document analysis is applied to extract important information (e.g., symptoms, duration of symptom occurrence, relevant medical history, etc.). The extracted information is stored on the server in JSON format or as a database record.

[0758] The server then uses a summary generator to generate a summary based on the extracted key information. A template-based summary generation algorithm is used. Specifically, a dynamic template engine (e.g., Jinja2) is used to match the extracted data with specific template sentences to create a concise and accurate summary. The generated summary sentences are stored in the server's database.

[0759] The generated summary is sent from the server to the doctor's terminal. The summary is displayed through a user interface, where the user (doctor) can confirm it. The user interface is interactive, allowing the doctor to easily make corrections to the summary. The corrected summary is sent back to the server and linked to the electronic medical record system, where it is saved as the final medical record.

[0760] Specific examples

[0761] For example, consider the case where a patient states:

[0762] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[0763] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text using a speech recognition API, resulting in the sentence, "For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold." The natural language processing engine analyzes this text data and extracts information such as "for a week," "every night," "cough," and "it gets worse when it's cold." The server then uses this information to generate a summary.

[0764] "I started coughing every night for the past week, and it gets worse when it's cold."

[0765] The doctor then checks it and saves it in the medical record.

[0766] Example prompts for generative AI models

[0767] Examples of prompts for using an AI model to generate summaries include:

[0768] "Summarize the following conversation and include the main symptom and its duration: 'For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold.'"

[0769] This system enables efficient and accurate creation of medical records by integrating the conversion of voice data into text data, extraction of important information, generation of summaries, provision and correction of the summaries to medical professionals, and storage of the summaries as medical records.

[0770] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0771] Step 1:

[0772] The terminal records conversations between patients and doctors in real time. The input is the audio of the conversation during the consultation, and the output is audio data temporarily stored in a local buffer. Specifically, the terminal uses a microphone to capture the audio and processes the audio data via a digital signal processor (DSP), resulting in clear audio data with little noise.

[0773] Step 2:

[0774] The device streams recorded audio data to the server. The input is audio data stored in a local buffer, and the output is audio data sent to the server in real time. Specifically, the device uses an Internet connection to stream data using a secure communication protocol such as HTTPS. This ensures that the audio data is securely sent to the server.

[0775] Step 3:

[0776] The server processes the received voice data using a speech recognition API and converts it into text data. The input is streamed voice data, and the output is converted text data. The server then calls a speech recognition API (such as Google Cloud Speech-to-Text or Amazon Transcribe) to convert the voice data into text. This allows the content of the conversation to be obtained as text information.

[0777] Step 4:

[0778] The server analyzes the text data using a natural language processing engine to extract important information. The input is the text data, and the output is the extracted information (e.g., symptoms, duration of onset, relevant medical history). Specifically, the server uses a natural language processing engine (e.g., spaCy, NLTK) to tokenize the text data and apply a key phrase extraction algorithm. This allows the necessary medical information to be accurately extracted from the text.

[0779] Step 5:

[0780] The server generates a summary based on the extracted information. The input is the extracted key information, and the output is the generated summary. The server uses a template-based summary generation algorithm and a dynamic template engine (e.g., Jinja2) to match the extracted data with the defined template, resulting in the generated summary.

[0781] Step 6:

[0782] The server sends the generated summary to the doctor's terminal. The input is the generated summary text, and the output is the summary text displayed on the doctor's terminal. Specifically, the server sends the summary text to the doctor's terminal and displays it through the user interface, allowing the doctor to immediately check the summary content.

[0783] Step 7:

[0784] User: The doctor reviews the summary and makes corrections if necessary. The input is the displayed summary, and the output is the corrected summary. Specifically, the doctor makes corrections to the summary using an interactive user interface and sends the corrections to the server, which then produces the final confirmed summary.

[0785] Step 8:

[0786] The server connects the revised summary to the electronic medical record system and stores it. The input is the revised summary, and the output is the medical record stored in the electronic medical record system. The server integrates the revised summary into the electronic medical record system and stores it as a medical record, enabling long-term management of medical data.

[0787] This provides a system that can efficiently generate and manage useful and accurate clinical data from a series of sessions.

[0788] (Application example 1)

[0789] 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."

[0790] In factory production management, there is a need to efficiently collect and convert voice instructions from workers and managers into text, extract important information, and generate summaries. However, conventional methods require a great deal of time and effort to convert voice instructions into text and extract information, resulting in a significant decrease in work efficiency. In particular, in large factories, delays and misunderstandings in work instructions are likely to occur, adversely affecting quality control and productivity.

[0791] 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.

[0792] In this invention, the server includes means for recording conversations with workers or managers as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important instructions from the text data, summary generation means for generating a summary based on the extracted important instructions, means for providing the generated summary to the worker or manager, and means for checking and correcting the provided summary, thereby enabling voice instructions to be quickly and accurately converted into text and important information to be efficiently extracted and summarized.

[0793] "Workers" refer to employees who perform production and maintenance work within a factory.

[0794] A "manager" is a person in a position to manage business operations and give work instructions within a factory.

[0795] "Voice data" refers to data that has been digitally recorded from the voices of workers and managers.

[0796] "Speech recognition means" refers to a technique or device for converting voice data into text data.

[0797] "Text data" refers to data that has been converted into character information based on voice data.

[0798] "Natural language processing means" refers to technology or devices that extract meaningful information from text data.

[0799] A "summary generator" is a technique or device that generates a concise summary based on the extracted information.

[0800] A "summary" is a document that briefly summarizes important instructions or information.

[0801] The "presentation means" refers to a technique or device that presents the generated summary to the worker or manager.

[0802] "Verification and correction means" refers to techniques or devices that allow an operator or manager to verify the provided summary and make corrections as necessary.

[0803] The present invention is a system for efficiently collecting voice instructions from workers and managers in a production management system for factory robots, converting the voice data into text data, and extracting important instruction content and generating summaries. Specific embodiments for carrying out the present invention will be described below.

[0804] First, when a worker or manager gives instructions or reports by voice in the factory, the device records this conversation as voice data in real time. The hardware used can be a microphone, smart glasses, or a head-mounted display. The recorded voice data is temporarily stored in a local buffer and simultaneously streamed to the server.

[0805] The server processes the received voice data. First, the server converts the voice data into text data using a speech recognition API. Typical speech recognition APIs include Google Speech-to-Text and IBM Watson Speech to Text. The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained generative AI model to extract important instructions (e.g., work content, deadlines, points to note, etc.) from the text data.

[0806] Specifically, the text data is tokenized and a key phrase extraction algorithm specialized for analyzing factory work instructions is applied. Natural language processing frameworks such as spaCy and NLTK are often used. After the important instruction content is extracted, the server generates a summary based on this information. A template-based summary generation algorithm is used to generate the summary. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the work content, deadlines, and important points.

[0807] The generated summary is sent from the server to the worker's or manager's terminal, where it can be viewed by the user through the terminal's user interface. The user, that is, the worker or manager, can review this summary and easily make corrections as necessary. An interface for this purpose is also provided. The corrected summary is finally saved in the production management system or work instruction system. This saving process is carried out in conjunction with the central management system.

[0808] As a specific example, consider the case where a worker states the following:

[0809] Worker: "Please complete the maintenance on Line 1 by tonight. Please be careful not to affect the operation of Line 2."

[0810] The terminal records this voice instruction and sends it as voice data to the server. The server converts the voice data into text, saying, "Please complete maintenance on Line 1 by tonight. Please take care not to affect the operation of Line 2." This text data is analyzed using a natural language processing engine, and information such as "tonight," "Line 1," "maintenance," and "operation of Line 2" is extracted. The server then uses this to generate a summary: "Complete maintenance on Line 1 by tonight. Take care to keep Line 2 running," and provides it to the user. The user then confirms this and saves it in the work instruction system.

[0811] Example prompts to input to a generative AI model:

[0812] Summarize the audio instructions below.

[0813] Voice command: "Please complete maintenance on Line 1 by this evening. Please be careful not to affect the operation of Line 2."

[0814] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0815] Step 1:

[0816] The terminal collects voice instructions from workers and managers using a microphone. The voice data is temporarily stored in a local buffer. The input is the voice of the worker or manager, and the output is voice data. The voice is captured using a microphone, converted into digital format, and stored.

[0817] Step 2:

[0818] The device streams locally stored audio data to the server in real time. The input is the audio data stored on the device, and the output is the audio data sent to the server. In this step, the audio data is sent to the server using an internet connection.

[0819] Step 3:

[0820] The server converts the received voice data into text data using a speech recognition API. The input is voice data and the output is text data. Specifically, APIs such as Google Speech-to-Text and IBM Watson Speech to Text are called to convert voice to text.

[0821] Step 4:

[0822] The server analyzes the text data using a natural language processing engine and extracts important instructions. The input is text data, and the output is important instructions (key phrases). Here, natural language processing frameworks such as spaCy and NLTK are used to tokenize the text and extract key phrases.

[0823] Step 5:

[0824] The server creates a summary based on the extracted key instructions using a template-based summary generation algorithm. The input is the key instructions, and the output is a summary. By embedding the extracted information in a template, a concise and accurate summary is generated.

[0825] Step 6:

[0826] The server sends the generated summary to the worker or administrator's terminal. The input is the summary text, and the output is the summary text displayed on the terminal. The sent summary text is delivered to the terminal via an Internet connection.

[0827] Step 7:

[0828] The terminal displays the summary to the worker or manager through a user interface. The input is the summary sent from the server, and the output is the displayed summary that the user can check. This step uses devices such as smart glasses or a head-mounted display.

[0829] Step 8:

[0830] The user, a worker or administrator, checks the provided summary and easily modifies it if necessary. The input is the displayed summary, and the output is the modified summary. The function to edit the summary is provided through the user interface.

[0831] Step 9:

[0832] The revised summary is finally saved in the production control system or work instruction system. The input is the revised summary, and the output is the saved summary. In this step, the summary is saved in cooperation with the database system.

[0833] 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.

[0834] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts the conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[0835] A patient and a doctor start talking in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to a server in real time.

[0836] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[0837] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[0838] Furthermore, the server uses an emotion engine that analyzes voice data to recognize the user's emotional state. This emotion engine analyzes the tone, rhythm, and speed of the voice and uses algorithms to detect the user's emotional state (e.g., relief, anxiety, anger, sadness, etc.). The emotion engine works in parallel with the speech-to-text conversion process using the speech recognition API.

[0839] The emotional state of the user recognized by the emotion engine is saved together with the analysis results of the natural language processing engine and reflected in the generation process of the summary generator, which generates a summary that includes information related to the user's emotional state in addition to the regular summary.

[0840] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[0841] Specific examples

[0842] For example, consider the case where a patient states:

[0843] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[0844] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. I seem anxious," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[0845] In this way, by combining the emotion engine, more detailed and accurate medical records can be generated that take into account the patient's emotional state, contributing to supporting doctors in their medical treatment.

[0846] The processing flow will be explained below.

[0847] Step 1: Start recording

[0848] User: Patient and doctor begin a consultation and conversation.

[0849] The device uses a microphone to record this conversation as audio data.

[0850] The recorded audio data is temporarily stored in a local buffer.

[0851] Step 2: Sending audio data

[0852] As the conversation progresses, the device streams audio data in real time to the server.

[0853] The streaming audio data is continuously received by the server.

[0854] Step 3: Speech recognition processing

[0855] The server passes the received voice data to a voice recognition API and converts it into text data.

[0856] The speech recognition API analyzes the audio data and generates a corresponding string.

[0857] The generated text data is temporarily stored for the next processing step.

[0858] Step 4: Emotion recognition processing

[0859] The server sends the voice data to the emotion engine to analyze the user's emotional state.

[0860] The emotion engine analyzes the tone, rhythm, and speed of speech to identify the user's emotional state (e.g., relief, anxiety, anger, sadness).

[0861] The emotional state recognized by the emotion engine is stored together with the text data.

[0862] Step 5: Analyzing the text data

[0863] The server uses a natural language processing engine to analyze the text data.

[0864] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[0865] The extracted information is organized as structured data.

[0866] Step 6: Generate a summary

[0867] The server generates a summary based on the structured data and the emotional state.

[0868] A template-based summary generation algorithm is used to create a summary sentence by applying the extracted information and emotional state to a template sentence.

[0869] The generated summary includes the patient's main symptoms, duration of symptoms, and emotional state.

[0870] Step 7: Provide a summary

[0871] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0872] The terminal receives the summary data and displays a screen that the physician can review and modify.

[0873] Step 8: Check and correct

[0874] User: The physician reviews the summary and makes corrections as needed through the interface.

[0875] The modified summary is then sent back to the server from the terminal.

[0876] Step 9: Save

[0877] The server receives the revised summary and stores it as the final data in the patient's medical record.

[0878] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[0879] This not only automatically generates the necessary medical information from the voice data, but also provides medical information that takes into account the patient's emotional state, improving the quality of medical care.

[0880] Example 2

[0881] 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."

[0882] In modern medical settings, there is a need to efficiently collect conversational information with patients and generate accurate medical records. However, simply converting voice data into text, extracting important information, and generating summaries does not provide medical support that takes into account the patient's emotional state. This can lead to a lack of understanding of the patient's psychological state, which could result in a decline in the quality of medical care.

[0883] 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.

[0884] In this invention, the server includes an emotion recognition means for analyzing the emotional state of the patient, a means for reflecting the analysis result of the emotion recognition means in the summary generation means, and a means for providing the generated summary, thereby enabling the generation of detailed and accurate medical records that also take into account the emotional state of the patient.

[0885] "Patient" means a person receiving medical services.

[0886] "Audio data" refers to recorded audio stored in digital format.

[0887] "Text data" refers to data obtained by converting voice data into character information.

[0888] "Speech recognition means" refers to technology or a system that converts voice data into text data.

[0889] "Natural language processing means" refers to technologies and systems that analyze and extract important information from text data.

[0890] A "summary generation means" is a technology or system that creates a summary based on the extracted information.

[0891] "Emotion recognition means" refers to technology or a system that analyzes voice data or text data and recognizes the user's emotional state.

[0892] A "summary" is a piece of text or data that briefly summarizes important information or key points.

[0893] "Means for providing" refers to the technology or system that presents the generated summary to doctors or other medical personnel.

[0894] "Verification and correction means" refers to techniques or systems that review the provided summary and correct it if necessary.

[0895] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts that conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions.

[0896] System configuration

[0897] 1. Terminal

[0898] The device is a device with audio capture capabilities for recording conversations. The device is equipped with a high-quality microphone, which records conversations in real time and temporarily stores them in a local buffer. The recorded audio data is streamed from the device to a server.

[0899] Specifically, the device used can be a smartphone or a dedicated recording device, and the start of recording can be triggered by, for example, pressing a button or using a voice command.

[0900] 2. Server

[0901] The server is composed of software with multiple functions, including voice recognition, natural language processing, and emotion recognition.

[0902] Specifically, it is composed as follows:

[0903] Speech recognition API: Use Google Cloud Speech-to-Text API or similar to convert voice data into text data.

[0904] Natural language processing engines: Use frameworks such as spaCy and NLTK to extract important information from text data.

[0905] Emotion Engine: Uses IBM Watson Tone Analyzer to analyze voice and text data and recognize the user's emotional state.

[0906] The server first receives the voice data and converts it into text data via a voice recognition API. It then analyzes the text data using a natural language processing engine to extract important information. At the same time, it analyzes the voice data using an emotion engine to recognize the user's emotional state. It then generates a summary based on the extracted important information and emotional state.

[0907] 3. User (Doctor)

[0908] The user, a physician, uses a terminal with an interface to review the generated summary and, if necessary, modify it. This interface displays the summary in a visually easy-to-understand format and allows editing.

[0909] The doctor can review the summary on the device and make any necessary corrections, after which the summary is sent back to the server and ultimately stored in the electronic medical record (EMR) system.

[0910] Specific examples

[0911] For example, consider the following situation where a patient states:

[0912] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[0913] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, he's had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "He started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. He seems anxious," and provides it to the doctor. The doctor then verifies this and saves it in the medical record.

[0914] This system will generate more detailed and accurate medical records that take into account the patient's emotional state, greatly assisting doctors in their medical practice.

[0915] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0916] Step 1:

[0917] When a conversation between a patient and a doctor begins, the device records the audio in real time using a microphone and temporarily stores it in a local buffer.

[0918] Specific operation: The patient and doctor start a conversation in the examination room, and the microphone on the device captures the audio. The device's button operation or voice command is used to trigger the start of recording.

[0919] Input: Voice of conversation between patient and doctor

[0920] Output: Locally stored audio data

[0921] Step 2:

[0922] The device streams the stored audio data to the server in real time.

[0923] Specific operation: The terminal maintains a stable network connection, divides the data stored in the buffer into certain packets, and delivers them to the server.

[0924] Input: Locally stored audio data

[0925] Output: Audio data streamed to the server

[0926] Step 3:

[0927] The server converts the received voice data into text data using a voice recognition API.

[0928] Specific operation: After receiving the voice data, the server sends the data to a speech recognition API and stores the returned text data in memory. For example, we use the Google Cloud Speech-to-Text API.

[0929] Input: Audio data streamed to the server

[0930] Output: Text data returned from the speech recognition API

[0931] Step 4:

[0932] The server analyzes the acquired text data using a natural language processing engine and extracts important information.

[0933] What it does: The server tokenizes the text data and applies algorithms (e.g., spaCy, NLTK) to extract key phrases such as symptoms, duration, and relevant medical history.

[0934] Input: Text data returned from the speech recognition API

[0935] Output: Key information extracted by the natural language processing engine

[0936] Step 5:

[0937] The server simultaneously inputs the voice data into an emotion recognition engine to recognize the patient's emotional state.

[0938] Specific operation: The server analyzes the tone, rhythm, speed, etc. of the voice data and identifies the emotional state (e.g., anxiety, relief, anger, sadness, etc.) using an emotion recognition algorithm (e.g., IBM Watson Tone Analyzer).

[0939] Input: Audio data streamed to the server

[0940] Output: Patient emotional state data from the emotion recognition engine

[0941] Step 6:

[0942] The server integrates the analyzed text data with the emotional state and generates a summary.

[0943] Specific operation: The server uses the extracted information and emotional state data to construct a summary sentence and stores it in memory.

[0944] Input: Important information extracted by a natural language processing engine, and patient emotional state data from an emotion recognition engine

[0945] Output: Generated summary

[0946] Step 7:

[0947] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[0948] Specific operation: The generated summary is sent to the terminal and the summary content is displayed on the interface, including a visual display and editing function for doctors to check the summary.

[0949] Input: Generated summary

[0950] Output: Summary displayed on the terminal

[0951] Step 8:

[0952] User: The physician reviews the provided summary and makes any necessary corrections.

[0953] Specific actions: Check the summary on the device, make corrections in edit mode, and press the confirm button to save the changes.

[0954] Input: Summary text displayed on the terminal

[0955] Output: revised summary

[0956] Step 9:

[0957] The server finally saves the verified and corrected summary as a medical record in the electronic medical record (EMR) system.

[0958] Specific operation: The server calls the API of the EMR system and stores the revised summary data in association with the patient information.

[0959] Input: revised summary

[0960] Output: Medical records stored in the electronic medical record

[0961] (Application example 2)

[0962] 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."

[0963] A problem with modern smartphone usage is the lack of real-time security assessment based on the user's emotional state and conversation content. This makes it difficult to properly predict potential cyber threats and security risks and take countermeasures. Furthermore, the lack of technology to generate security alerts that reflect the user's unstable emotional state prevents users from using digital devices with peace of mind. To solve this problem, a fast and accurate security alert generation system that utilizes the user's conversation information and emotional state is needed.

[0964] 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.

[0965] In this invention, the server includes means for recording conversations with patients as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important information from the text data, summary generation means for generating a summary based on the extracted important information, means for providing the generated summary to a person in charge, means for confirming and correcting the provided summary, emotion analysis means for analyzing the emotional state, and means for creating and notifying a security alert based on the analysis results, thereby enabling real-time security evaluation and alert notification based on the user's conversation information and emotional state.

[0966] "Patient" refers to an individual receiving medical care or treatment.

[0967] "Audio data" refers to a data file that digitally records audio.

[0968] "Speech recognition means" refers to technology or devices for analyzing voice data and converting it into text data.

[0969] "Text data" refers to information expressed in text, typically stored in digital form.

[0970] "Natural language processing means" refers to technologies and devices for analyzing text data, understanding its meaning, and extracting important information.

[0971] "Significant information" refers to data that is particularly useful in relation to a particular purpose or use.

[0972] "Summary generator" refers to a technique or device that summarizes the extracted important information in a concise, easy-to-understand format.

[0973] "Contact Person" refers to the individual or group responsible for handling the generated summary and / or security assessment results.

[0974] "Emotion analysis means" refers to technology or devices for analyzing and determining a user's emotional state from voice or text data.

[0975] "Security Alert" refers to notifications intended to warn users of potential cyber threats or security risks.

[0976] "Means for creating and notifying security alerts" refers to technologies and devices for issuing security warnings to users based on the results of sentiment analysis and natural language processing means.

[0977] This invention is a system that analyzes a user's conversation information and emotional state, performs security assessments in real time, and generates and notifies appropriate security alerts.

[0978] First, the smartphone device records the conversation with the user as audio data. The audio data collected using the microphone is streamed to a server in real time. This communication uses the smartphone's internet connection. The collection and communication of this audio data uses general technology that is independent of the smartphone's operating system.

[0979] The server analyzes the received voice data. First, it uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) as a speech recognition method to convert the voice data into text data. This converted text data is then analyzed by a natural language processing method (e.g., spaCy). The natural language processing engine uses a pre-trained model to extract important information from the text data. Important information includes the user's activity, the period during which the activity occurred, and related historical information.

[0980] Next, the server uses emotion analysis to analyze the user's emotional state from the voice data. This emotion analysis takes into account the tone, rhythm, and speed of the voice. For example, OpenAI's emotion analysis model is used for emotion analysis. Emotional states such as relief, anxiety, anger, and sadness are determined.

[0981] The analyzed text data and emotional state are combined into a single summary using a summary generator. The summary is generated using a template-based summary generation algorithm. The generated summary is provided to the human resource, who can review the summary and make any necessary revisions.

[0982] Furthermore, the server creates and notifies the user of security alerts based on the analysis results. The security alert evaluates potential cyber threats based on the user's emotional state and conversation content, and provides a warning accordingly. For example, if the emotional state is anxiety and the conversation content is "Could it be a virus?", the system will immediately notify the user with a recommendation to scan for a virus.

[0983] As a concrete example, consider the following user conversation:

[0984] "My phone has been running slow lately. Could this be due to a virus?" (Tone sounds concerned)

[0985] The smartphone records this conversation and sends it to the server. The server converts the voice into text, saying, "My smartphone has been running slowly lately. Could this be because I have a virus?" The text data is analyzed using a natural language processing engine, and information such as "My smartphone is running slowly" and "Possibly a virus" is extracted. In parallel, an emotion analysis tool analyzes the voice data and recognizes the user's emotional state as "worried." Based on this information, the server generates a summary: "My smartphone is running slowly, and I'm worried that it might have a virus." A person in charge checks this summary, and then the user is notified with a security alert, such as "We recommend a virus scan."

[0986] An example prompt for a generative AI model might look like this:

[0987] Input user conversation data as text: My smartphone has been running slow lately. Could it be because of a virus?

[0988] Enter the corresponding emotional state: Worry

[0989] Integrated results: Generate virus scanning recommendations based on conversation concerns.

[0990] The recommendation it generates is: Virus scanning and establishing security measures are recommended.

[0991] As a result, the present invention can propose security measures in real time based on the user's conversation information and emotional state, and provide an environment in which the user can use the device with peace of mind.

[0992] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0993] Step 1:

[0994] The smartphone device records the user's conversation. The user's conversation is acquired as audio data as input. Specifically, the smartphone's microphone is used to record the conversation in digital form. The audio data is generated as output.

[0995] Step 2:

[0996] The smartphone device streams audio data to a server in real time. Recorded audio data is used as input. Specifically, the smartphone's communication module is used to send the data to the server via the Internet. The server receives the audio data as output.

[0997] Step 3:

[0998] The server uses a speech recognition tool to convert the voice data into text data. The voice data streamed to the server is used as input. Specifically, a speech recognition API (e.g., Google Cloud Speech-to-Text API) is used to analyze and convert the voice data. Text data is generated as output.

[0999] Step 4:

[1000] The server uses natural language processing means to analyze the text data and extract important information. The text data generated by speech recognition is used as input. Specifically, a natural language processing engine (e.g., spaCy) is used to analyze the text data and extract important information (such as user activity, time period, and historical information). The important information is extracted as output.

[1001] Step 5:

[1002] The server uses emotion analysis means to analyze the user's emotional state from the voice data. The voice data is used as input. Specifically, an algorithm based on an emotion analysis model (e.g., OpenAI's emotion analysis model) is used to analyze the emotional state (relief, anxiety, anger, sadness, etc.) from the tone, rhythm, and speed of the voice. The analyzed emotional state is output.

[1003] Step 6:

[1004] The server uses a summary generation means to generate a summary based on the extracted key information and the analyzed emotional state. The key information and the emotional state are used as input. Specifically, a template-based summary generation algorithm is used to generate the summary. The summary is generated as output.

[1005] Step 7:

[1006] The server provides the generated summary to the person in charge. The generated summary is used as input. Specifically, the server sends the summary to the person in charge's terminal. As output, the summary is ready for the person in charge to check.

[1007] Step 8:

[1008] The person in charge checks the summary and makes corrections if necessary. The provided summary is used as input. Specifically, the person in charge checks the provided summary on their terminal and edits it using the correction interface. The corrected summary is generated as output.

[1009] Step 9:

[1010] The server creates a security alert based on the analysis results and notifies the user. The summary and sentiment analysis results are used as input. Specifically, an alert message is generated based on the analysis results and sent to the user's smartphone. The user receives the security alert as output.

[1011] 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.

[1012] 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.

[1013] 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.

[1014] [Fourth embodiment]

[1015] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1016] 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.

[1017] 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).

[1018] 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.

[1019] 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.

[1020] 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).

[1021] 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.

[1022] 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.

[1023] 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.

[1024] 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.

[1025] 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.

[1026] 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.

[1027] 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."

[1028] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention will now be described.

[1029] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to the server in real time.

[1030] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[1031] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[1032] After extracting the key information, the server uses this information to generate a summary. This is done using a template-based summary generation algorithm. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the patient's main symptoms, when the symptoms began, and their medical history.

[1033] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[1034] Specific examples

[1035] For example, consider the case where a patient states:

[1036] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[1037] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." The text data is analyzed using a natural language processing engine, and information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold" is extracted. The server then uses this information to generate a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[1038] In this way, the present invention enables efficient and accurate generation of medical records, reduces the burden on doctors, and significantly reduces patient waiting times.

[1039] The processing flow will be explained below.

[1040] Step 1: Start recording

[1041] User: Patient and doctor begin a consultation and conversation.

[1042] The device uses a microphone to record this conversation as audio data.

[1043] The recorded audio data is temporarily stored in a local buffer.

[1044] Step 2: Sending audio data

[1045] As the conversation progresses, the device streams audio data in real time to the server.

[1046] The streaming audio data is continuously received by the server.

[1047] Step 3: Speech recognition processing

[1048] The server passes the received voice data to a voice recognition API and converts it into text data.

[1049] The speech recognition API analyzes the audio data and generates a corresponding string.

[1050] The generated text data is temporarily stored for the next processing step.

[1051] Step 4: Analyzing the text data

[1052] The server uses a natural language processing engine to analyze the text data.

[1053] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[1054] The extracted information is organized as structured data.

[1055] Step 5: Generate a summary

[1056] The server generates a summary based on the structured data.

[1057] A template-based summary generation algorithm is used to create a summary by applying the extracted information to a template sentence.

[1058] The generated summary includes the patient's main symptoms and the duration of symptoms.

[1059] Step 6: Provide a summary

[1060] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[1061] The terminal receives the summary data and displays a screen that the physician can review and modify.

[1062] Step 7: Check and correct

[1063] User: The physician reviews the summary and makes corrections as needed through the interface.

[1064] The modified summary is then sent back to the server from the terminal.

[1065] Step 8: Save

[1066] The server receives the revised summary and stores it as the final data in the patient's medical record.

[1067] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[1068] This will automatically generate the necessary medical information from the voice data, reducing the burden on doctors and making the medical treatment process more efficient.

[1069] Example 1

[1070] 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."

[1071] In modern medical settings, it is important for medical professionals to quickly and accurately record conversations with patients and efficiently create medical records. However, manual recording is time-consuming and labor-intensive, and there is a risk of errors. In addition, there is a lack of systems that can comprehensively convert voice data into text data and extract and summarize important information. Therefore, a solution to this problem is needed.

[1072] 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.

[1073] In this invention, the server includes a device that records conversations with patients as voice data, a voice recognition device that converts the voice data into text data, and a natural language processing device that extracts important information from the text data. This allows for the creation of quick and accurate medical records by including a device that streams the voice data over the Internet, a device that saves the extracted information in JSON format or as a database record, a device that uses a dynamic template engine to generate summaries, and a device that saves the summaries in an electronic medical record system. This reduces the burden on medical professionals and shortens patient waiting times.

[1074] A "device for recording conversations with patients as audio data" is a device for capturing and recording conversations between patients and doctors as audio data using a microphone or the like.

[1075] A "voice recognition device that converts voice data into text data" is software or hardware that analyzes recorded voice data and converts it into corresponding text data.

[1076] A "natural language processing device that extracts important information from text data" is a device that includes algorithms and models for analyzing text data and automatically extracting necessary information such as a patient's symptoms and the duration of onset.

[1077] The "summary generation device that generates a summary based on extracted important information" refers to software or hardware that organizes the extracted important information and creates a summary using a template or the like.

[1078] The "apparatus for providing the generated summary to the medical professional" refers to a device or interface for displaying or linking the generated summary so that the medical professional can review it.

[1079] A "device for reviewing and correcting provided summaries" is an interface or software that allows a healthcare professional to review the generated summaries and make corrections as necessary.

[1080] A "device for streaming audio data over the Internet" is a communication device for transmitting recorded audio data to a server in real time.

[1081] The "device that stores extracted information in JSON format or as database records" is a software component that stores extracted important information as structured data.

[1082] An "apparatus using a dynamic template engine for generating summaries" is a software tool for dynamically applying extracted data to templates to generate summaries.

[1083] The "device that links and saves summaries in an electronic medical record system" refers to an interface or software that saves the generated summaries in an electronic medical record system and links them with the necessary medical information.

[1084] The present invention provides a system for efficiently collecting conversational information with patients, converting the conversational data into text data, extracting important information, and generating a summary. Specific embodiments for carrying out the present invention are described in detail below.

[1085] First, a patient and a doctor start a conversation in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and processes it with a digital signal processor (DSP) to obtain high-quality audio data. The recorded audio data is temporarily stored in a local buffer and then streamed to a server via the Internet. A secure communication protocol such as HTTPS is used to securely transmit the audio data.

[1086] The server processes the received voice data. Specifically, it converts the voice data into text data using a voice recognition device. Common cloud services such as Google Cloud Speech-to-Text and Amazon Transcribe can be used as the voice recognition API. The converted text data is temporarily stored in the server's database.

[1087] Next, a natural language processing system on the server analyzes this text data. A common natural language processing framework, such as spaCy or NLTK, is used as the natural language processing engine. The text data is first tokenized, and a key phrase extraction algorithm specialized for medical document analysis is applied to extract important information (e.g., symptoms, duration of symptom occurrence, relevant medical history, etc.). The extracted information is stored on the server in JSON format or as a database record.

[1088] The server then uses a summary generator to generate a summary based on the extracted key information. A template-based summary generation algorithm is used. Specifically, a dynamic template engine (e.g., Jinja2) is used to match the extracted data with specific template sentences to create a concise and accurate summary. The generated summary sentences are stored in the server's database.

[1089] The generated summary is sent from the server to the doctor's terminal. The summary is displayed through a user interface, where the user (doctor) can confirm it. The user interface is interactive, allowing the doctor to easily make corrections to the summary. The corrected summary is sent back to the server and linked to the electronic medical record system, where it is saved as the final medical record.

[1090] Specific examples

[1091] For example, consider the case where a patient states:

[1092] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold."

[1093] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text using a speech recognition API, resulting in the sentence, "For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold." The natural language processing engine analyzes this text data and extracts information such as "for a week," "every night," "cough," and "it gets worse when it's cold." The server then uses this information to generate a summary.

[1094] "I started coughing every night for the past week, and it gets worse when it's cold."

[1095] The doctor then checks it and saves it in the medical record.

[1096] Example prompts for generative AI models

[1097] Examples of prompts for using an AI model to generate summaries include:

[1098] "Summarize the following conversation and include the main symptom and its duration: 'For the past week or so, I've had a persistent cough every night, and it gets worse when it's cold.'"

[1099] This system enables efficient and accurate creation of medical records by integrating the conversion of voice data into text data, extraction of important information, generation of summaries, provision and correction of the summaries to medical professionals, and storage of the summaries as medical records.

[1100] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1101] Step 1:

[1102] The terminal records conversations between patients and doctors in real time. The input is the audio of the conversation during the consultation, and the output is audio data temporarily stored in a local buffer. Specifically, the terminal uses a microphone to capture the audio and processes the audio data via a digital signal processor (DSP), resulting in clear audio data with little noise.

[1103] Step 2:

[1104] The device streams recorded audio data to the server. The input is audio data stored in a local buffer, and the output is audio data sent to the server in real time. Specifically, the device uses an Internet connection to stream data using a secure communication protocol such as HTTPS. This ensures that the audio data is securely sent to the server.

[1105] Step 3:

[1106] The server processes the received voice data using a speech recognition API and converts it into text data. The input is streamed voice data, and the output is converted text data. The server then calls a speech recognition API (such as Google Cloud Speech-to-Text or Amazon Transcribe) to convert the voice data into text. This allows the content of the conversation to be obtained as text information.

[1107] Step 4:

[1108] The server analyzes the text data using a natural language processing engine to extract important information. The input is the text data, and the output is the extracted information (e.g., symptoms, duration of onset, relevant medical history). Specifically, the server uses a natural language processing engine (e.g., spaCy, NLTK) to tokenize the text data and apply a key phrase extraction algorithm. This allows the necessary medical information to be accurately extracted from the text.

[1109] Step 5:

[1110] The server generates a summary based on the extracted information. The input is the extracted key information, and the output is the generated summary. The server uses a template-based summary generation algorithm and a dynamic template engine (e.g., Jinja2) to match the extracted data with the defined template, resulting in the generated summary.

[1111] Step 6:

[1112] The server sends the generated summary to the doctor's terminal. The input is the generated summary text, and the output is the summary text displayed on the doctor's terminal. Specifically, the server sends the summary text to the doctor's terminal and displays it through the user interface, allowing the doctor to immediately check the summary content.

[1113] Step 7:

[1114] User: The doctor reviews the summary and makes corrections if necessary. The input is the displayed summary, and the output is the corrected summary. Specifically, the doctor makes corrections to the summary using an interactive user interface and sends the corrections to the server, which then produces the final confirmed summary.

[1115] Step 8:

[1116] The server connects the revised summary to the electronic medical record system and stores it. The input is the revised summary, and the output is the medical record stored in the electronic medical record system. The server integrates the revised summary into the electronic medical record system and stores it as a medical record, enabling long-term management of medical data.

[1117] This provides a system that can efficiently generate and manage useful and accurate clinical data from a series of sessions.

[1118] (Application example 1)

[1119] 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."

[1120] In factory production management, there is a need to efficiently collect and convert voice instructions from workers and managers into text, extract important information, and generate summaries. However, conventional methods require a great deal of time and effort to convert voice instructions into text and extract information, resulting in a significant decrease in work efficiency. In particular, in large factories, delays and misunderstandings in work instructions are likely to occur, adversely affecting quality control and productivity.

[1121] 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.

[1122] In this invention, the server includes means for recording conversations with workers or managers as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important instructions from the text data, summary generation means for generating a summary based on the extracted important instructions, means for providing the generated summary to the worker or manager, and means for checking and correcting the provided summary, thereby enabling voice instructions to be quickly and accurately converted into text and important information to be efficiently extracted and summarized.

[1123] "Workers" refer to employees who perform production and maintenance work within a factory.

[1124] A "manager" is a person in a position to manage business operations and give work instructions within a factory.

[1125] "Voice data" refers to data that has been digitally recorded from the voices of workers and managers.

[1126] "Speech recognition means" refers to a technique or device for converting voice data into text data.

[1127] "Text data" refers to data that has been converted into character information based on voice data.

[1128] "Natural language processing means" refers to technology or devices that extract meaningful information from text data.

[1129] A "summary generator" is a technique or device that generates a concise summary based on the extracted information.

[1130] A "summary" is a document that briefly summarizes important instructions or information.

[1131] The "presentation means" refers to a technique or device that presents the generated summary to the worker or manager.

[1132] "Verification and correction means" refers to techniques or devices that allow an operator or manager to verify the provided summary and make corrections as necessary.

[1133] The present invention is a system for efficiently collecting voice instructions from workers and managers in a production management system for factory robots, converting the voice data into text data, and extracting important instruction content and generating summaries. Specific embodiments for carrying out the present invention will be described below.

[1134] First, when a worker or manager gives instructions or reports by voice in the factory, the device records this conversation as voice data in real time. The hardware used can be a microphone, smart glasses, or a head-mounted display. The recorded voice data is temporarily stored in a local buffer and simultaneously streamed to the server.

[1135] The server processes the received voice data. First, the server converts the voice data into text data using a speech recognition API. Typical speech recognition APIs include Google Speech-to-Text and IBM Watson Speech to Text. The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained generative AI model to extract important instructions (e.g., work content, deadlines, points to note, etc.) from the text data.

[1136] Specifically, the text data is tokenized and a key phrase extraction algorithm specialized for analyzing factory work instructions is applied. Natural language processing frameworks such as spaCy and NLTK are often used. After the important instruction content is extracted, the server generates a summary based on this information. A template-based summary generation algorithm is used to generate the summary. Specifically, the extracted data is matched to specific template sentences to create a concise and accurate summary. This summary includes information such as the work content, deadlines, and important points.

[1137] The generated summary is sent from the server to the worker's or manager's terminal, where it can be viewed by the user through the terminal's user interface. The user, that is, the worker or manager, can review this summary and easily make corrections as necessary. An interface for this purpose is also provided. The corrected summary is finally saved in the production management system or work instruction system. This saving process is carried out in conjunction with the central management system.

[1138] As a specific example, consider the case where a worker states the following:

[1139] Worker: "Please complete the maintenance on Line 1 by tonight. Please be careful not to affect the operation of Line 2."

[1140] The terminal records this voice instruction and sends it as voice data to the server. The server converts the voice data into text, saying, "Please complete maintenance on Line 1 by tonight. Please take care not to affect the operation of Line 2." This text data is analyzed using a natural language processing engine, and information such as "tonight," "Line 1," "maintenance," and "operation of Line 2" is extracted. The server then uses this to generate a summary: "Complete maintenance on Line 1 by tonight. Take care to keep Line 2 running," and provides it to the user. The user then confirms this and saves it in the work instruction system.

[1141] Example prompts to input to a generative AI model:

[1142] Summarize the audio instructions below.

[1143] Voice command: "Please complete maintenance on Line 1 by this evening. Please be careful not to affect the operation of Line 2."

[1144] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1145] Step 1:

[1146] The terminal collects voice instructions from workers and managers using a microphone. The voice data is temporarily stored in a local buffer. The input is the voice of the worker or manager, and the output is voice data. The voice is captured using a microphone, converted into digital format, and stored.

[1147] Step 2:

[1148] The device streams locally stored audio data to the server in real time. The input is the audio data stored on the device, and the output is the audio data sent to the server. In this step, the audio data is sent to the server using an internet connection.

[1149] Step 3:

[1150] The server converts the received voice data into text data using a speech recognition API. The input is voice data and the output is text data. Specifically, APIs such as Google Speech-to-Text and IBM Watson Speech to Text are called to convert voice to text.

[1151] Step 4:

[1152] The server analyzes the text data using a natural language processing engine and extracts important instructions. The input is text data, and the output is important instructions (key phrases). Here, natural language processing frameworks such as spaCy and NLTK are used to tokenize the text and extract key phrases.

[1153] Step 5:

[1154] The server creates a summary based on the extracted key instructions using a template-based summary generation algorithm. The input is the key instructions, and the output is a summary. By embedding the extracted information in a template, a concise and accurate summary is generated.

[1155] Step 6:

[1156] The server sends the generated summary to the worker or administrator's terminal. The input is the summary text, and the output is the summary text displayed on the terminal. The sent summary text is delivered to the terminal via an Internet connection.

[1157] Step 7:

[1158] The terminal displays the summary to the worker or manager through a user interface. The input is the summary sent from the server, and the output is the displayed summary that the user can check. This step uses devices such as smart glasses or a head-mounted display.

[1159] Step 8:

[1160] The user, a worker or administrator, checks the provided summary and easily modifies it if necessary. The input is the displayed summary, and the output is the modified summary. The function to edit the summary is provided through the user interface.

[1161] Step 9:

[1162] The revised summary is finally saved in the production control system or work instruction system. The input is the revised summary, and the output is the saved summary. In this step, the summary is saved in cooperation with the database system.

[1163] 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.

[1164] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts the conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions. Specific embodiments for carrying out the present invention will be described below.

[1165] A patient and a doctor start talking in the examination room. The device records this conversation as audio data in real time. The device uses a microphone to capture the audio and temporarily stores the audio data in a local buffer. The stored audio data is then streamed to a server in real time.

[1166] The server processes the received voice data and converts it into text using a speech recognition API. For example, a common speech recognition API can be a cloud service or an on-premise solution for converting speech to text.

[1167] The converted text data is then analyzed by a natural language processing engine on the server. This natural language processing engine uses a pre-trained model to extract important information from the text data (e.g., symptoms, the duration of symptoms, relevant medical history, etc.). Specifically, it tokenizes the text data and applies a key phrase extraction algorithm specialized for medical document analysis. This is achieved by using a common natural language processing framework (e.g., spaCy or NLTK).

[1168] Furthermore, the server uses an emotion engine that analyzes voice data to recognize the user's emotional state. This emotion engine analyzes the tone, rhythm, and speed of the voice and uses algorithms to detect the user's emotional state (e.g., relief, anxiety, anger, sadness, etc.). The emotion engine works in parallel with the speech-to-text conversion process using the speech recognition API.

[1169] The emotional state of the user recognized by the emotion engine is saved together with the analysis results of the natural language processing engine and reflected in the generation process of the summary generator, which generates a summary that includes information related to the user's emotional state in addition to the regular summary.

[1170] The generated summary is sent from the server to the doctor's terminal, where it can be viewed by the doctor through the terminal's user interface. User: The doctor can review this summary and easily make corrections as needed. An interface for this purpose is also provided. The corrected summary is finally saved on the server as the patient's medical record. This saving process is carried out in conjunction with the electronic medical record (EMR) system.

[1171] Specific examples

[1172] For example, consider the case where a patient states:

[1173] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[1174] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, I've had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "I started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. I seem anxious," which is provided to the user / doctor. The doctor then verifies this and saves it in the medical record.

[1175] In this way, by combining the emotion engine, more detailed and accurate medical records can be generated that take into account the patient's emotional state, contributing to supporting doctors in their medical treatment.

[1176] The processing flow will be explained below.

[1177] Step 1: Start recording

[1178] User: Patient and doctor begin a consultation and conversation.

[1179] The device uses a microphone to record this conversation as audio data.

[1180] The recorded audio data is temporarily stored in a local buffer.

[1181] Step 2: Sending audio data

[1182] As the conversation progresses, the device streams audio data in real time to the server.

[1183] The streaming audio data is continuously received by the server.

[1184] Step 3: Speech recognition processing

[1185] The server passes the received voice data to a voice recognition API and converts it into text data.

[1186] The speech recognition API analyzes the audio data and generates a corresponding string.

[1187] The generated text data is temporarily stored for the next processing step.

[1188] Step 4: Emotion recognition processing

[1189] The server sends the voice data to the emotion engine to analyze the user's emotional state.

[1190] The emotion engine analyzes the tone, rhythm, and speed of speech to identify the user's emotional state (e.g., relief, anxiety, anger, sadness).

[1191] The emotional state recognized by the emotion engine is stored together with the text data.

[1192] Step 5: Analyzing the text data

[1193] The server uses a natural language processing engine to analyze the text data.

[1194] The NLP engine tokenizes the text data and extracts medically relevant key phrases and important information (e.g., symptoms, duration, relevant medical history).

[1195] The extracted information is organized as structured data.

[1196] Step 6: Generate a summary

[1197] The server generates a summary based on the structured data and the emotional state.

[1198] A template-based summary generation algorithm is used to create a summary sentence by applying the extracted information and emotional state to a template sentence.

[1199] The generated summary includes the patient's main symptoms, duration of symptoms, and emotional state.

[1200] Step 7: Provide a summary

[1201] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[1202] The terminal receives the summary data and displays a screen that the physician can review and modify.

[1203] Step 8: Check and correct

[1204] User: The physician reviews the summary and makes corrections as needed through the interface.

[1205] The modified summary is then sent back to the server from the terminal.

[1206] Step 9: Save

[1207] The server receives the revised summary and stores it as the final data in the patient's medical record.

[1208] The stored data is automatically linked to the electronic medical record (EMR) system and used as medical records.

[1209] This not only automatically generates the necessary medical information from the voice data, but also provides medical information that takes into account the patient's emotional state, improving the quality of medical care.

[1210] Example 2

[1211] 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."

[1212] In modern medical settings, there is a need to efficiently collect conversational information with patients and generate accurate medical records. However, simply converting voice data into text, extracting important information, and generating summaries does not provide medical support that takes into account the patient's emotional state. This can lead to a lack of understanding of the patient's psychological state, which could result in a decline in the quality of medical care.

[1213] 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.

[1214] In this invention, the server includes an emotion recognition means for analyzing the emotional state of the patient, a means for reflecting the analysis result of the emotion recognition means in the summary generation means, and a means for providing the generated summary, thereby enabling the generation of detailed and accurate medical records that also take into account the emotional state of the patient.

[1215] "Patient" means a person receiving medical services.

[1216] "Audio data" refers to recorded audio stored in digital format.

[1217] "Text data" refers to data obtained by converting voice data into character information.

[1218] "Speech recognition means" refers to technology or a system that converts voice data into text data.

[1219] "Natural language processing means" refers to technologies and systems that analyze and extract important information from text data.

[1220] A "summary generation means" is a technology or system that creates a summary based on the extracted information.

[1221] "Emotion recognition means" refers to technology or a system that analyzes voice data or text data and recognizes the user's emotional state.

[1222] A "summary" is a piece of text or data that briefly summarizes important information or key points.

[1223] "Means for providing" refers to the technology or system that presents the generated summary to doctors or other medical personnel.

[1224] "Verification and correction means" refers to techniques or systems that review the provided summary and correct it if necessary.

[1225] The present invention improves the accuracy of medical information by combining a system that efficiently collects conversational information with patients, converts that conversational data into text data, extracts important information, and generates summaries, with an emotion engine that recognizes the user's emotions.

[1226] System configuration

[1227] 1. Terminal

[1228] The device is a device with audio capture capabilities for recording conversations. The device is equipped with a high-quality microphone, which records conversations in real time and temporarily stores them in a local buffer. The recorded audio data is streamed from the device to a server.

[1229] Specifically, the device used can be a smartphone or a dedicated recording device, and the start of recording can be triggered by, for example, pressing a button or using a voice command.

[1230] 2. Server

[1231] The server is composed of software with multiple functions, including voice recognition, natural language processing, and emotion recognition.

[1232] Specifically, it is composed as follows:

[1233] Speech recognition API: Use Google Cloud Speech-to-Text API or similar to convert voice data into text data.

[1234] Natural language processing engines: Use frameworks such as spaCy and NLTK to extract important information from text data.

[1235] Emotion Engine: Uses IBM Watson Tone Analyzer to analyze voice and text data and recognize the user's emotional state.

[1236] The server first receives the voice data and converts it into text data via a voice recognition API. It then analyzes the text data using a natural language processing engine to extract important information. At the same time, it analyzes the voice data using an emotion engine to recognize the user's emotional state. It then generates a summary based on the extracted important information and emotional state.

[1237] 3. User (Doctor)

[1238] The user, a physician, uses a terminal with an interface to review the generated summary and, if necessary, modify it. This interface displays the summary in a visually easy-to-understand format and allows editing.

[1239] The doctor can review the summary on the device and make any necessary corrections, after which the summary is sent back to the server and ultimately stored in the electronic medical record (EMR) system.

[1240] Specific examples

[1241] For example, consider the following situation where a patient states:

[1242] User: Patient: "For the past week or so, I've had a constant cough every night, and it gets worse when it's cold." (Speaking in a weak, anxious voice)

[1243] The device records this conversation and sends it as audio data to the server. The server converts the audio data into text, stating, "For the past week or so, he's had a persistent cough in the middle of the night every day, and it gets worse when it's cold." This text data is analyzed using a natural language processing engine, which extracts information such as "for a week," "every day in the middle of the night," "cough," and "it gets worse when it's cold." In parallel, the emotion engine analyzes the audio data and recognizes the patient's emotional state as "anxiety." Based on this, the server generates a summary: "He started coughing in the middle of the night every day for the past week, and it gets worse when it's cold. He seems anxious," and provides it to the doctor. The doctor then verifies this and saves it in the medical record.

[1244] This system will generate more detailed and accurate medical records that take into account the patient's emotional state, greatly assisting doctors in their medical practice.

[1245] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1246] Step 1:

[1247] When a conversation between a patient and a doctor begins, the device records the audio in real time using a microphone and temporarily stores it in a local buffer.

[1248] Specific operation: The patient and doctor start a conversation in the examination room, and the microphone on the device captures the audio. The device's button operation or voice command is used to trigger the start of recording.

[1249] Input: Voice of conversation between patient and doctor

[1250] Output: Locally stored audio data

[1251] Step 2:

[1252] The device streams the stored audio data to the server in real time.

[1253] Specific operation: The terminal maintains a stable network connection, divides the data stored in the buffer into certain packets, and delivers them to the server.

[1254] Input: Locally stored audio data

[1255] Output: Audio data streamed to the server

[1256] Step 3:

[1257] The server converts the received voice data into text data using a voice recognition API.

[1258] Specific operation: After receiving the voice data, the server sends the data to a speech recognition API and stores the returned text data in memory. For example, we use the Google Cloud Speech-to-Text API.

[1259] Input: Audio data streamed to the server

[1260] Output: Text data returned from the speech recognition API

[1261] Step 4:

[1262] The server analyzes the acquired text data using a natural language processing engine and extracts important information.

[1263] What it does: The server tokenizes the text data and applies algorithms (e.g., spaCy, NLTK) to extract key phrases such as symptoms, duration, and relevant medical history.

[1264] Input: Text data returned from the speech recognition API

[1265] Output: Key information extracted by the natural language processing engine

[1266] Step 5:

[1267] The server simultaneously inputs the voice data into an emotion recognition engine to recognize the patient's emotional state.

[1268] Specific operation: The server analyzes the tone, rhythm, speed, etc. of the voice data and identifies the emotional state (e.g., anxiety, relief, anger, sadness, etc.) using an emotion recognition algorithm (e.g., IBM Watson Tone Analyzer).

[1269] Input: Audio data streamed to the server

[1270] Output: Patient emotional state data from the emotion recognition engine

[1271] Step 6:

[1272] The server integrates the analyzed text data with the emotional state and generates a summary.

[1273] Specific operation: The server uses the extracted information and emotional state data to construct a summary sentence and stores it in memory.

[1274] Input: Important information extracted by a natural language processing engine, and patient emotional state data from an emotion recognition engine

[1275] Output: Generated summary

[1276] Step 7:

[1277] The server sends the generated summary to the terminal so that the doctor can check it through the user interface.

[1278] Specific operation: The generated summary is sent to the terminal and the summary content is displayed on the interface, including a visual display and editing function for doctors to check the summary.

[1279] Input: Generated summary

[1280] Output: Summary displayed on the terminal

[1281] Step 8:

[1282] User: The physician reviews the provided summary and makes any necessary corrections.

[1283] Specific actions: Check the summary on the device, make corrections in edit mode, and press the confirm button to save the changes.

[1284] Input: Summary text displayed on the terminal

[1285] Output: revised summary

[1286] Step 9:

[1287] The server finally saves the verified and corrected summary as a medical record in the electronic medical record (EMR) system.

[1288] Specific operation: The server calls the API of the EMR system and stores the revised summary data in association with the patient information.

[1289] Input: revised summary

[1290] Output: Medical records stored in the electronic medical record

[1291] (Application example 2)

[1292] 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 robot 414 will be referred to as a "terminal."

[1293] A problem with modern smartphone usage is the lack of real-time security assessment based on the user's emotional state and conversation content. This makes it difficult to properly predict potential cyber threats and security risks and take countermeasures. Furthermore, the lack of technology to generate security alerts that reflect the user's unstable emotional state prevents users from using digital devices with peace of mind. To solve this problem, a fast and accurate security alert generation system that utilizes the user's conversation information and emotional state is needed.

[1294] 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.

[1295] In this invention, the server includes means for recording conversations with patients as voice data, speech recognition means for converting the voice data into text data, natural language processing means for extracting important information from the text data, summary generation means for generating a summary based on the extracted important information, means for providing the generated summary to a person in charge, means for confirming and correcting the provided summary, emotion analysis means for analyzing the emotional state, and means for creating and notifying a security alert based on the analysis results, thereby enabling real-time security evaluation and alert notification based on the user's conversation information and emotional state.

[1296] "Patient" refers to an individual receiving medical care or treatment.

[1297] "Audio data" refers to a data file that digitally records audio.

[1298] "Speech recognition means" refers to technology or devices for analyzing voice data and converting it into text data.

[1299] "Text data" refers to information expressed in text, typically stored in digital form.

[1300] "Natural language processing means" refers to technologies and devices for analyzing text data, understanding its meaning, and extracting important information.

[1301] "Significant information" refers to data that is particularly useful in relation to a particular purpose or use.

[1302] "Summary generator" refers to a technique or device that summarizes the extracted important information in a concise, easy-to-understand format.

[1303] "Contact Person" refers to the individual or group responsible for handling the generated summary and / or security assessment results.

[1304] "Emotion analysis means" refers to technology or devices for analyzing and determining a user's emotional state from voice or text data.

[1305] "Security Alert" refers to notifications intended to warn users of potential cyber threats or security risks.

[1306] "Means for creating and notifying security alerts" refers to technologies and devices for issuing security warnings to users based on the results of sentiment analysis and natural language processing means.

[1307] This invention is a system that analyzes a user's conversation information and emotional state, performs security assessments in real time, and generates and notifies appropriate security alerts.

[1308] First, the smartphone device records the conversation with the user as audio data. The audio data collected using the microphone is streamed to a server in real time. This communication uses the smartphone's internet connection. The collection and communication of this audio data uses general technology that is independent of the smartphone's operating system.

[1309] The server analyzes the received voice data. First, it uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) as a speech recognition method to convert the voice data into text data. This converted text data is then analyzed by a natural language processing method (e.g., spaCy). The natural language processing engine uses a pre-trained model to extract important information from the text data. Important information includes the user's activity, the period during which the activity occurred, and related historical information.

[1310] Next, the server uses emotion analysis to analyze the user's emotional state from the voice data. This emotion analysis takes into account the tone, rhythm, and speed of the voice. For example, OpenAI's emotion analysis model is used for emotion analysis. Emotional states such as relief, anxiety, anger, and sadness are determined.

[1311] The analyzed text data and emotional state are combined into a single summary using a summary generator. The summary is generated using a template-based summary generation algorithm. The generated summary is provided to the human resource, who can review the summary and make any necessary revisions.

[1312] Furthermore, the server creates and notifies the user of security alerts based on the analysis results. The security alert evaluates potential cyber threats based on the user's emotional state and conversation content, and provides a warning accordingly. For example, if the emotional state is anxiety and the conversation content is "Could it be a virus?", the system will immediately notify the user with a recommendation to scan for a virus.

[1313] As a concrete example, consider the following user conversation:

[1314] "My phone has been running slow lately. Could this be due to a virus?" (Tone sounds concerned)

[1315] The smartphone records this conversation and sends it to the server. The server converts the voice into text, saying, "My smartphone has been running slowly lately. Could this be because I have a virus?" The text data is analyzed using a natural language processing engine, and information such as "My smartphone is running slowly" and "Possibly a virus" is extracted. In parallel, an emotion analysis tool analyzes the voice data and recognizes the user's emotional state as "worried." Based on this information, the server generates a summary: "My smartphone is running slowly, and I'm worried that it might have a virus." A person in charge checks this summary, and then the user is notified with a security alert, such as "We recommend a virus scan."

[1316] An example prompt for a generative AI model might look like this:

[1317] Input user conversation data as text: My smartphone has been running slow lately. Could it be because of a virus?

[1318] Enter the corresponding emotional state: Worry

[1319] Integrated results: Generate virus scanning recommendations based on conversation concerns.

[1320] The recommendation it generates is: Virus scanning and establishing security measures are recommended.

[1321] As a result, the present invention can propose security measures in real time based on the user's conversation information and emotional state, and provide an environment in which the user can use the device with peace of mind.

[1322] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1323] Step 1:

[1324] The smartphone device records the user's conversation. The user's conversation is acquired as audio data as input. Specifically, the smartphone's microphone is used to record the conversation in digital form. The audio data is generated as output.

[1325] Step 2:

[1326] The smartphone device streams audio data to a server in real time. Recorded audio data is used as input. Specifically, the smartphone's communication module is used to send the data to the server via the Internet. The server receives the audio data as output.

[1327] Step 3:

[1328] The server uses a speech recognition tool to convert the voice data into text data. The voice data streamed to the server is used as input. Specifically, a speech recognition API (e.g., Google Cloud Speech-to-Text API) is used to analyze and convert the voice data. Text data is generated as output.

[1329] Step 4:

[1330] The server uses natural language processing means to analyze the text data and extract important information. The text data generated by speech recognition is used as input. Specifically, a natural language processing engine (e.g., spaCy) is used to analyze the text data and extract important information (such as user activity, time period, and historical information). The important information is extracted as output.

[1331] Step 5:

[1332] The server uses emotion analysis means to analyze the user's emotional state from the voice data. The voice data is used as input. Specifically, an algorithm based on an emotion analysis model (e.g., OpenAI's emotion analysis model) is used to analyze the emotional state (relief, anxiety, anger, sadness, etc.) from the tone, rhythm, and speed of the voice. The analyzed emotional state is output.

[1333] Step 6:

[1334] The server uses a summary generation means to generate a summary based on the extracted key information and the analyzed emotional state. The key information and the emotional state are used as input. Specifically, a template-based summary generation algorithm is used to generate the summary. The summary is generated as output.

[1335] Step 7:

[1336] The server provides the generated summary to the person in charge. The generated summary is used as input. Specifically, the server sends the summary to the person in charge's terminal. As output, the summary is ready for the person in charge to check.

[1337] Step 8:

[1338] The person in charge checks the summary and makes corrections if necessary. The provided summary is used as input. Specifically, the person in charge checks the provided summary on their terminal and edits it using the correction interface. The corrected summary is generated as output.

[1339] Step 9:

[1340] The server creates a security alert based on the analysis results and notifies the user. The summary and sentiment analysis results are used as input. Specifically, an alert message is generated based on the analysis results and sent to the user's smartphone. The user receives the security alert as output.

[1341] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice 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 voice data.

[1342] 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.

[1343] 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 robot 414.

[1344] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1345] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1346] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1347] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1348] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1349] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1350] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1351] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1352] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1353] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1354] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1355] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1356] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1357] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1358] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1359] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1360] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1361] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1362] The following is further disclosed regarding the above embodiment.

[1363] (Claim 1)

[1364] a means for recording conversations with patients as audio data;

[1365] a speech recognition means for converting speech data into text data;

[1366] natural language processing means for extracting important information from text data;

[1367] a summary generation means for generating a summary based on the extracted important information;

[1368] a means for providing the generated summary to a physician or other person;

[1369] The system includes a means for reviewing and correcting the provided summary.

[1370] (Claim 2)

[1371] 10. The system of claim 1, wherein the extracted important information includes the patient's symptoms, the duration of the symptoms, and related medical history information.

[1372] (Claim 3)

[1373] 10. The system of claim 1, wherein the summary generating means generates the summary using a template-based summary generation algorithm.

[1374] "Example 1"

[1375] (Claim 1)

[1376] A device for recording conversations with patients as audio data;

[1377] a voice recognition device that converts voice data into text data;

[1378] A natural language processing device that extracts important information from text data;

[1379] a summary generation device that generates a summary based on the extracted important information;

[1380] a device for providing the generated summary to a healthcare professional;

[1381] a device for verifying and correcting the provided summary;

[1382] a device for streaming audio data over the Internet;

[1383] A device that stores the extracted information in JSON format or as a database record;

[1384] an apparatus that uses a dynamic template engine for generating summaries;

[1385] A device that links and saves summaries in an electronic medical record system

[1386] A system including:

[1387] (Claim 2)

[1388] 2. The system of claim 1, wherein the extracted important information includes the patient's symptoms, the duration of the symptoms, relevant medical history information, and conditions for the worsening of the symptoms.

[1389] (Claim 3)

[1390] 10. The system of claim 1, wherein the summary generator generates the summary using a template-based summary generation algorithm and a dynamic template engine.

[1391] "Application Example 1"

[1392] (Claim 1)

[1393] a means for recording a conversation with a worker or a manager as audio data;

[1394] a speech recognition means for converting speech data into text data;

[1395] natural language processing means for extracting important instruction contents from text data;

[1396] a summary generation means for generating a summary based on the extracted important instruction content;

[1397] a means for providing the generated summary to a worker or manager;

[1398] The system includes a means for reviewing and correcting the provided summary.

[1399] (Claim 2)

[1400] 2. The system according to claim 1, wherein the extracted important instructions include work content, deadlines, and points to note.

[1401] (Claim 3)

[1402] 10. The system of claim 1, wherein the summary generating means generates the summary using a template-based summary generation algorithm.

[1403] "Example 2: Combining Emotion Engines"

[1404] (Claim 1)

[1405] a means for recording conversations with patients as audio data;

[1406] a speech recognition means for converting speech data into text data;

[1407] natural language processing means for extracting important information from text data;

[1408] a summary generation means for generating a summary based on the extracted important information;

[1409] an emotion recognition means for analyzing an emotional state;

[1410] A means for reflecting the analysis result of the emotion recognition means in the summary generation means;

[1411] a means for providing the generated summary;

[1412] The system includes a means for reviewing and correcting the provided summary.

[1413] (Claim 2)

[1414] 10. The system of claim 1, wherein the extracted important information includes the patient's symptoms, the duration of the symptoms, and related medical history information.

[1415] (Claim 3)

[1416] 2. The system according to claim 1, wherein the summary generating means generates a summary including the analysis result of the emotion recognition means.

[1417] (Claim 4)

[1418] 10. The system of claim 1, wherein the summary generating means generates the summary using a template-based summary generation algorithm.

[1419] "Application example 2 when combining emotion engines"

[1420] (Claim 1)

[1421] a means for recording conversations with patients as audio data;

[1422] a speech recognition means for converting speech data into text data;

[1423] natural language processing means for extracting important information from text data;

[1424] a summary generation means for generating a summary based on the extracted important information;

[1425] a means for providing the generated summary to a person in charge;

[1426] a means to review and amend the summary provided;

[1427] emotion analysis means for analyzing an emotional state;

[1428] A system that includes a means to create and notify security alerts based on the analysis results.

[1429] (Claim 2)

[1430] 2. The system of claim 1, wherein the extracted important information includes user activities, time periods during which the activities occurred, and related historical information.

[1431] (Claim 3)

[1432] 10. The system of claim 1, wherein the summary generating means generates the summary using a template-based summary generation algorithm. [Explanation of symbols]

[1433] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for recording conversations with patients as audio data; a speech recognition means for converting speech data into text data; natural language processing means for extracting important information from text data; a summary generation means for generating a summary based on the extracted important information; a means for providing the generated summary to a physician or other person; The system includes a means for reviewing and correcting the provided summary.

2. 10. The system of claim 1, wherein the extracted important information includes the patient's symptoms, the duration of the symptoms, and related medical history information.

3. 10. The system of claim 1, wherein the summary generating means generates the summary using a template-based summary generation algorithm.

Images