Program, information processing method, and information processing apparatus.

Abstract

To provide a program or similar tool that can identify healthcare professionals to whom tasks in medical facilities can be assigned. [Solution] The program, relating to one aspect, acquires tasks at a medical facility and, based on the acquired tasks and the output from a model that uses the medical professional information of multiple medical professionals, causes the computer to perform a process to identify the medical professional to whom the tasks will be assigned. This makes it possible to identify the medical professional to whom the tasks at the medical facility will be assigned.

Description

Technical Field

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

Background Art

[0008] [Figure 1] This is an explanatory diagram showing an overview of the nursing work assignment system. [Figure 2] This is a block diagram showing an example server configuration. [Figure 3] This is an explanatory diagram showing an example of the record layout for the nurse database and patient database. [Figure 4] This is an explanatory diagram showing an example of a record layout in the task database. [Figure 5] This is a block diagram showing an example configuration of a nurse terminal. [Figure 6] This is an explanatory diagram illustrating the process for identifying the nurse to whom a task will be assigned. [Figure 7] This is an explanatory diagram illustrating the process of assigning tasks to nurses. [Figure 8] This flowchart shows the process for assigning tasks to nurses. [Figure 9] This is an explanatory diagram showing an example of the record layout of the task database in variation 3. [Figure 10] This is an explanatory diagram illustrating the outline of the nurse identification process using a task assignment model. [Figure 11] This is an explanatory diagram illustrating the process of assigning tasks to nurses in Embodiment 2. [Figure 12] This flowchart shows the processing procedure for assigning tasks to nurses in Embodiment 2. [Modes for carrying out the invention]

[0009] The present invention will be described in detail below with reference to the drawings illustrating its embodiments.

[0010] (Embodiment 1) Embodiment 1 relates to a form of identifying medical staff who assign tasks (operations) in a medical facility. FIG. 1 is an explanatory diagram showing an overview of a nursing task allocation system. The system of this embodiment includes an information processing apparatus 1 and a plurality of information processing terminals 2, and each apparatus transmits and receives information via a network N such as the Internet.

[0011] The information processing apparatus 1 is an information processing apparatus that performs processing, storage, transmission, and reception of various information. The information processing apparatus 1 is, for example, a server apparatus, a personal computer, or a general-purpose tablet PC (personal computer), etc. In this embodiment, the information processing apparatus 1 is assumed to be a server apparatus, and hereinafter it will be read as server 1 for simplicity.

[0012] The information processing terminal 2 is a terminal device for medical staff that receives and displays a task list (ToDo list) in which task plans (for example, medicine administration, examination, or observation, etc.) for each patient in a medical facility are displayed in time series. In the following, an example where the medical staff is a nurse will be described, but it can be similarly applied to other types of medical staff such as doctors or pharmacists.

[0013] The information processing terminal 2 is an information processing device such as a personal computer terminal, a tablet, a smartphone, a mobile phone, or a wearable device such as a smartwatch. In the following, for simplicity, the information processing terminal 2 will be read as nurse terminal 2.

[0014] In a medical facility, in order to appropriately handle the determination of the business order of a plurality of tasks, the priority order of tasks related to the lives of patients, time-specified examinations, treatments, toilet assistance, or tasks according to the requests of patients, an administrator (for example, a head nurse) needs to perform appropriate staffing. However, there is a problem that inexperienced administrators cannot perform optimal staffing.

[0015] A medical facility is a place or organization for providing medical services to patients, and includes a hospital, a clinic, a dispensary pharmacy, an emergency medical center, or a health examination center, etc.

[0016] In order to solve such problems, in this embodiment, a nurse who uses artificial intelligence (AI) to assign tasks in a medical facility can be identified. As a result, the workload of the administrator can be reduced.

[0017] The server 1 according to this embodiment acquires tasks in a medical facility. The server 1 uses a model to identify a nurse to whom the acquired tasks are to be assigned. Note that the tasks and the model will be described later.

[0018] FIG. 2 is a block diagram showing a configuration example of the server 1. The server 1 includes a control unit 11, a storage unit 12, a communication unit 13, a reading unit 14, and a mass storage unit 15. Each component is connected by a bus B.

[0019] The control unit 11 includes an arithmetic processing device such as a CPU (Central Processing Unit), MPU (Micro-Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), or a quantum processor. By reading and executing the control program 1P (program product) stored in the storage unit 12, the control unit 11 performs various information processing and control processing related to the server 1.

[0020] Note that the control program 1P can be deployed to be executed on a single computer, or at one site, or distributed over a plurality of sites and executed on a plurality of computers interconnected by a communication network.

[0021] In Figure 2, the control unit 11 is described as a single processor, but it may be a multi-processor system. Furthermore, the control unit 11 may perform various information processing or control processing on the same processor within the server 1, or it may perform these processes on different processors within the server 1.

[0022] The memory unit 12 includes memory elements such as RAM (Random Access Memory) and ROM (Read Only Memory), and stores the control program 1P or data necessary for the control unit 11 to execute processing. The memory unit 12 also temporarily stores data necessary for the control unit 11 to execute arithmetic processing. The communication unit 13 is a communication module for performing communication-related processing, and sends and receives information with the nurse terminal 2 and other devices via the network N.

[0023] The reading unit 14 reads a portable storage medium 1a, including a CD (Compact Disc)-ROM or DVD (Digital Versatile Disc)-ROM. The control unit 11 may read the control program 1P from the portable storage medium 1a via the reading unit 14 and store it in the large-capacity storage unit 15. Alternatively, the control unit 11 may download the control program 1P from another computer via a network N or the like and store it in the large-capacity storage unit 15. Furthermore, the control unit 11 may also read the control program 1P from the semiconductor memory 1b.

[0024] The large-capacity storage unit 15 includes a recording medium such as an HDD (Hard disk drive) or an SSD (Solid State Drive). The large-capacity storage unit 15 includes a task assignment model 151, a nurse database 152, a patient database 153, and a task database 154.

[0025] The task assignment model 151 in this embodiment is a language generation model constructed by pre-training on a large-scale text data set. As the task assignment model 151, for example, large-scale language models (LLMs) such as Transformer, ALBERT (A Lite BERT), GPT (Generative Pre-trained Transformer)-2, GPT-3, GPT-4, LLaVA (Large Language and Vision Assistant), MiniGPT-4, or BERT (Bidirectional Encoder Representations from Transformers) can be used.

[0026] Nurse DB152 stores information about nurses. Patient DB153 stores information about patients. Task DB154 stores tasks assigned to nurses in the healthcare facility.

[0027] In this embodiment, the storage unit 12 and the large-capacity storage unit 15 may be configured as a single storage device. Furthermore, the large-capacity storage unit 15 may be composed of multiple storage devices. Moreover, the large-capacity storage unit 15 may be an external storage device connected to the server 1.

[0028] Server 1 may perform various information processing and control processes as a single computer, or it may be performed in a distributed manner across multiple computers. Furthermore, Server 1 may be implemented using multiple virtual machines located within a single server, or it may be implemented using a cloud server.

[0029] Figure 3 is an explanatory diagram showing an example of the record layout for Nurse DB152 and Patient DB153. Nurse DB152 includes columns for Nurse ID, Workdays, Work Type, Work Hours, Subject, Years of Experience, and Skills. The Nurse ID column stores the unique ID of each nurse to identify them. The Workdays column stores the days of the week on which nurses work. The Work Type column stores the type of work, including "day shift" and "night shift," etc.

[0030] The "Working Hours" column records the working hours. The "Subject" column records the specialty subjects the nurses specialize in. These specialties include, for example, emergency medicine, internal medicine, surgery, cardiology, or psychosomatic medicine. The "Years of Experience" column records the total number of years of experience as a nurse. The "Skills" column records the names of the nurses' skills.

[0031] Skills include, for example, "Ladder 1," "Ladder 2," "Ladder 3," "Emergency Response Certification," or "Cardiology Certification." For example, "Ladder 1" is a certification that proves basic nursing skills. "Ladder 2" is a certification that proves intermediate nursing skills. "Ladder 3" is a certification that proves advanced nursing skills. "Emergency Response Certification" is a certification that proves a nurse has specialized knowledge and skills in the field of emergency medicine. "Cardiology Certification" is a certification that proves specialized knowledge and skills in the nursing of patients with heart disease.

[0032] Patient DB153 includes columns for Patient ID, Name, Gender, Diagnosis, and Specialty. The Patient ID column stores a uniquely identifiable patient ID to identify each patient. The Name column stores the patient's name. The Gender column stores the patient's gender. The Diagnosis column stores the diagnosis of the patient's illness. The Specialty column stores the specialty (internal medicine, surgery, or cardiology, etc.) related to the diagnosis or treatment.

[0033] Figure 4 is an explanatory diagram showing an example of the record layout of Task DB154. Task DB154 includes columns for Task ID, Date, Time, Task Description, and Nurse ID. The Task ID column stores the uniquely identified task ID to identify each task. The Date column stores the date on which the task is performed.

[0034] The Time column stores the time the task was performed. The Task Content column stores the content of the task (e.g., taking temperature, performing an electrocardiogram, or handling an emergency admission). The Nurse ID column stores the Nurse ID of the nurse who assigned the task.

[0035] The storage configurations described above for each database are merely examples; other storage configurations are also acceptable as long as the relationships between the data are maintained.

[0036] Figure 5 is a block diagram showing an example configuration of the nurse terminal 2. The nurse terminal 2 includes a control unit 21, a storage unit 22, a communication unit 23, an input unit 24, and a display unit 25.

[0037] The control unit 21 includes a processing unit such as a CPU or MPU, and performs various information processing and control processing related to the nurse terminal 2 by reading and executing the control program 2P (program product) stored in the storage unit 22.

[0038] In Figure 5, the control unit 21 is described as a single processor, but it may be a multi-processor. Furthermore, the control unit 21 may perform various information processing or control processing using the same processor within the nurse terminal 2, or it may perform these processes using different processors within the nurse terminal 2.

[0039] The storage unit 22 includes memory elements such as RAM or ROM, and stores the control program 2P or data necessary for the control unit 21 to execute processing. The storage unit 22 also temporarily stores data necessary for the control unit 21 to execute arithmetic processing.

[0040] The communication unit 23 is a communication module for performing communication-related processing and sends and receives information with the server 1, etc., via the network N. The input unit 24 may be a keyboard, mouse, or a touch panel integrated with the display unit 25. The display unit 25 is a liquid crystal display or an organic EL (electroluminescence) display, etc., and displays various information according to the instructions of the control unit 21.

[0041] Figure 6 is an explanatory diagram illustrating the process for identifying the nurse to whom a task will be assigned. Server 1 retrieves tasks at the medical facility from Task DB 154 based on the target date (e.g., today). A task includes the examinations and task details instructed by a physician, and time information for performing each task. Task details include patient information (e.g., patient ID), examinations such as taking the patient's temperature, electrocardiogram, or blood test, emergency response cases (e.g., emergency hospitalization or response to a sudden change in condition), toilet assistance, or meal service. Patient information may also include, for example, the diagnosis and medical specialty (e.g., "internal medicine").

[0042] Tasks are managed daily in the task database 154. Tasks such as examinations, temperature checks, or meal service for patients are registered (stored) in the task database 154 in advance. For example, a physician's terminal accepts input of examination items and examination dates and times for patients. The physician's terminal sends a task to server 1, including the accepted examination items (e.g., electrocardiogram) and examination date and time, associated with the patient ID. Server 1 assigns a task ID to the task sent from the physician's terminal. Server 1 stores the examination date and time (date and time) and task details (e.g., "Electrocardiogram (Patient: P-002)") in the task database 154, associated with the assigned task ID.

[0043] Alternatively, Nurse Terminal 2 for the head nurse accepts input of tasks such as rounds (patrolling wards or floors, etc.) or tasks such as taking a patient's temperature, serving meals, or administering medication. Nurse Terminal 2 sends the received tasks to Server 1, associating them with the patient ID. Server 1 assigns a task ID to the tasks sent from Nurse Terminal 2. Server 1 stores the date and time (for example, August 22nd, 8:00) and the task details (for example, "Temperature check (Patient: P-001)") in Task DB 154, associating them with the assigned task ID.

[0044] Server 1 generates a prompt to be given to the task assignment model 151 based on the acquired task (task content, task execution time, etc.). The task assignment model 151 is a language model that uses tasks and nursing information of multiple nurses, and is used as a program module that is part of artificial intelligence software.

[0045] The task assignment model 151 is trained using a large amount of training data. The training data includes information on multiple tasks and nurses in a medical facility, as well as the nurses to whom each task is assigned. The training process for the task assignment model 151 may be performed by server 1 or by an external information processing device. Alternatively, a pre-trained language model may be used without performing the training process.

[0046] For example, Server 1 performs a conversion process to transform the acquired training data into a format (tokens) that the task assignment model 151 can process. That is, Server 1 performs training on a tokenizer that divides the acquired text data into tokens. Server 1 uses the divided tokens to train the task assignment model 151 for a specific task (for example, answering questions).

[0047] In addition, the task assignment model 151 may be configured so that, instead of being stored in the large-capacity storage unit 15, the server 1 accesses and reads the data from an external language processing server or language processing platform.

[0048] The task assignment model 151 in this embodiment is a pre-built language model (language generation model) that takes as input a prompt including the acquired task, an instruction to identify the nurse to whom the task should be assigned from among several candidate nurses, and nurse information of several nurses to be referenced, and outputs the nurse to whom the task should be assigned.

[0049] An order is a directive to identify the nurse to whom a task will be assigned, based on the nursing information, task content, and execution time of multiple nurses. The nursing information includes, for example, the nurse ID, working days (e.g., "Monday to Saturday"), type of work (e.g., "day shift" or "night shift"), working hours (e.g., "8:00 to 16:00"), specialties (e.g., "emergency response" and "internal medicine"), years of experience (e.g., 2 years), or skills (e.g., "ladder 1").

[0050] An instruction might read, for example, "For multiple tasks, refer to the nursing profiles of multiple candidate nurses, and, considering each nurse's skills and work information (work days, work types, and working hours), identify which nurse will be assigned to each task from among the multiple nurses."

[0051] A prompt is an instruction or input statement created in a format understandable to the task assignment model 151 and given as input to the task assignment model 151. The task assignment model 151 interprets the input prompt and outputs an appropriate response (e.g., assign a nurse).

[0052] For example, the task assignment model 151 divides the prompt into tokens so that it can be converted into a format that it can process. The task assignment model 151 performs contextual understanding by calculating the relationship between each token in the prompt and other tokens.

[0053] The task assignment model 151 performs response generation processing to prompts based on linguistic knowledge obtained through pre-training and fine-tuning. For example, the task assignment model 151 uses generation methods such as greedy decoding, beam search, or sampling to select the optimal token. The task assignment model 151 decodes the selected token to convert it back into text format and generates a response assigning a nurse the task to.

[0054] Server 1 generates a prompt that includes the acquired task, an instruction to identify the nurse to whom the task should be assigned from among several candidate nurses, nurse information of several nurses to be referenced, and the output format of the identification result.

[0055] As shown in the diagram, one example of a generated prompt is: "For the following tasks, refer to the nursing information of multiple candidate nurses stored in Nurse DB152, and consider each nurse's assigned subject, years of experience, skills, and work information (work days, work type, work hours) to identify which nurse will be assigned to each task from among the multiple nurses." The identification results should be output in a format that shows the correspondence between the task and the assigned nurse. For example, the output format should be "Task ID: xxxx ⇒ Assigned Nurse: xxxx". Task list (To-Do list): 1. Task ID: T-0001 Date: August 22 Time: 8:00 Task details: taking temperature Patient information: Patient ID:P-001 Medical department: Internal medicine Diagnosis: Diabetes 2. Task ID: T-0002 Date: August 22 Time: 8:00 Task details: taking temperature Patient information: Patient ID:P-002 Medical Department: Cardiology Diagnosis: Heart disease It could also be "xxxxxxx".

[0056] Server 1 inputs the generated prompt into the task assignment model 151 and outputs identification results that identify the nurse to whom each task will be assigned.

[0057] As shown in the diagram, as an example, the specific result output was: Task ID: T-0001 ⇒ Nurse in Charge: N-001 Task ID: T-0002 ⇒ Nurse in Charge: N-002 It could also be "xxxxxxx".

[0058] Server 1 transmits (outputs) the identification results output from the task assignment model 151 to the nurse's terminal 2 of the nurse's administrator (e.g., head nurse). The identification results include each task ID and information about the nurse to whom each task was assigned (e.g., nurse ID or name).

[0059] The administrator's nurse terminal 2 receives the identification result sent from server 1. The administrator's nurse terminal 2 displays the received identification result on the screen. If the administrator's nurse terminal 2 has obtained approval for each task from the administrator, it assigns each task to the appropriate nurse. The approval process will be described later.

[0060] Figure 7 is an explanatory diagram illustrating the process of assigning tasks to nurses. This screen includes multiple first task list display fields 11a, a second task list display field 11b, a specific result display field 11e, and an assign button 11f.

[0061] The first task list display area 11a is a display area for showing a task list in chronological order, displaying one or more assigned tasks for each nurse. The tasks include the task content and execution time. The second task list display area 11b is a display area for showing a task list that includes one or more unassigned tasks.

[0062] Each of the first task list display area 11a and the second task list display area 11b includes a nurse display area 11c and a task icon 11d. The nurse display area 11c of the first task list display area 11a is a display area that displays information about the nurse (e.g., nurse ID or name). The nurse display area 11c of the second task list display area 11b displays text indicating that no task has been assigned (e.g., "Assignee undecided").

[0063] Task icon 11d is an icon that displays assigned tasks (task details and execution time, etc.) or allows for the acceptance of an unassigned task to be assigned to a nurse. In this embodiment, an example has been described in which the text displayed on task icon 11d is the task details and execution time, but this is not the only example. For example, the text displayed on task icon 11d may also include the progress status of the task (e.g., not started, in progress, or completed).

[0064] In this embodiment, an example was described in which the task icon 11d is an icon, but it is not limited to this and may be a button, an image, or the like.

[0065] The Identification Result Display Area 11e is a display area for displaying the identification result of the nurse assigned to the target task. The Identification Result Display Area 11e includes the Recommended Nurse Display Area 11g and the Candidate Nurse Display Area 11h. The Recommended Nurse Display Area 11g is a display area for displaying the nurse information (e.g., nurse ID) of the nurse identified by the task assignment model 151. The Candidate Nurse Display Area 11h is a display area for displaying the nurse information of a candidate nurse different from the nurse identified by the task assignment model 151.

[0066] The assign button 11f is a button for accepting administrator approval for a task.

[0067] Server 1 retrieves tasks already assigned to each nurse, as well as unassigned tasks registered in Task DB 154, based on the current date (for example, August 22nd). Server 1 then sends the retrieved assigned and unassigned tasks to the administrator nurse's terminal 2.

[0068] Nurse terminal 2 receives assigned and unassigned tasks sent from server 1. Nurse terminal 2 displays the received assigned tasks for each nurse in the first task list display area 11a. As shown in the figure, for nurse ID "N-001", the tasks "16:00 Round", "18:00 Meal Service", and "19:00 Temperature Check (P-001)" are displayed on the task icons 11d in the first task list display area 11a. "P-001" is the patient ID of the patient targeted for the 19:00 temperature check task. Other patient information, such as the patient's name, may also be displayed on the task icon 11d.

[0069] Nurse terminal 2 displays unassigned tasks registered in task DB 154 in the second task list display area 11b. As shown in the figure, the tasks "Patient (P-001) Toilet Assistance" and "Patient (P-005) Emergency Hospitalization Response" are displayed as task icons 11d in the second task list display area 11b.

[0070] When Nurse Terminal 2 receives a touch (click) operation on the task icon 11d in the second task list display field 11b corresponding to the task to be assigned to the nurse, it sends the target task, including the task ID, patient information, and task details, to Server 1. Server 1 receives the target task from Nurse Terminal 2. Server 1 generates a prompt that includes the received target task, an instruction to identify the nurse to whom the target task should be assigned from among several candidate nurses, and nurse information of several nurses to be referenced.

[0071] Server 1 inputs the generated prompt into the task assignment model 151 and outputs an identification result that identifies the nurse to whom the target task will be assigned. Server 1 sends the identification result to the nurse terminal 2. Nurse terminal 2 receives the identification result sent from Server 1. Nurse terminal 2 displays the received identification result in the recommended nurse display field 11g within the identification result display field 11e. As shown in the figure, the identification result "Recommended Nurse: N-002" is displayed in the recommended nurse display field 11g.

[0072] Nurse terminal 2 retrieves nurses from among multiple nurses who are different from the nurse (N-002) identified by the task assignment model 151. Nurse terminal 2 displays the retrieved nurses in the candidate nurse display field 11h. As shown in the figure, the nurses with nurse IDs "N-001" and "N-003" retrieved from among multiple nurses (nurse IDs: "N-001", "N-002", and "N-003") are displayed in their respective candidate nurse display fields 11h.

[0073] Furthermore, the nurse terminal 2 may display nurse information such as the skills, years of experience, and continuous working hours of each nurse (recommended nurse or candidate nurse) in the identified results display field 11e. This allows the administrator to refer to this information when deciding which nurse to assign a task to.

[0074] When the nurse terminal 2 receives a touch operation on the recommended nurse display field 11g or the candidate nurse display field 11h within the specified result display field 11e, it obtains the nurse ID of the selected nurse. The selected nurse is either a nurse identified by the task assignment model 151 (for example, a nurse with nurse ID "N-002") or a different nurse (for example, a nurse with nurse ID "N-001" or "N-003"). As shown in the figure, the selected nurse is a nurse identified by the task assignment model 151.

[0075] When Nurse Terminal 2 receives a touch operation of the assign button 11f by the administrator, it sends the received approval information to Server 1. The approval information includes the task ID of the target task and the nurse ID of the selected nurse, etc.

[0076] Server 1 receives approval information sent from nurse terminal 2. Server 1 assigns the target task to the selected nurse according to the received approval information. Server 1 stores the nurse ID of the nurse who was assigned the target task in task DB 154, associating it with the task ID of the target task. Server 1 resends the task list, including the newly assigned target task, to nurse terminal 2.

[0077] Nurse terminal 2 receives the task list retransmitted from server 1 and redisplays the received task list in the first task list display area 11a and the second task list display area 11b. As shown in the figure, the task icon 11d for "Patient (P-005) Emergency Admission Response" displayed in the second task list display area 11b moves to the first task list display area 11a corresponding to the nurse (nurse ID: N-002) to whom the task is assigned and is displayed there.

[0078] Figure 8 is a flowchart showing the processing procedure for assigning tasks to nurses. The control unit 11 of server 1 retrieves a task list for the medical facility from the task DB 154 of the large-capacity storage unit 15 based on the target date (for example, today) (step S101). The task list includes a task list that displays single or multiple tasks assigned to each nurse in chronological order, and a task list that includes single or multiple unassigned tasks.

[0079] The control unit 11 transmits the acquired task list to the nurse administrator's terminal 2 via the communication unit 13 (step S102). The control unit 21 of the nurse terminal 2 receives the task list transmitted from the server 1 via the communication unit 23 (step S201). The control unit 21 displays the received task list using the display unit 25 (step S202).

[0080] The control unit 21 receives a task to be assigned to a nurse from the unassigned task list via the input unit 24 (step S203). The control unit 21 transmits the received task to the server 1 via the communication unit 13 (step S204). The task includes the task ID, task details (e.g., "Emergency Hospitalization Response"), and execution time.

[0081] The control unit 11 of server 1 receives the target task transmitted from the nurse terminal 2 via the communication unit 13 (step S103). Based on the received target task, the control unit 11 generates a prompt to be given to the task assignment model 151 (step S104). The prompt includes the target task, an instruction to identify the nurse to whom the task should be assigned from among several candidate nurses, and nurse information of several nurses to be referenced.

[0082] The control unit 11 inputs the generated prompt to the task assignment model 151 (step S105) and outputs an identification result that identifies the nurse to whom the target task will be assigned (step S106). The control unit 11 transmits the identification result output from the task assignment model 151 to the nurse terminal 2 via the communication unit 13 (step S107). The identification result includes the target task ID and information about the nurse to whom the task was assigned (e.g., nurse ID or name).

[0083] The control unit 21 of the nurse terminal 2 receives the specific result transmitted from the server 1 via the communication unit 23 (step S205). The nurse terminal 2 displays the received specific result via the display unit 25 (step S206). The control unit 21 receives approval for the target task from the administrator via the input unit 24 (for example, by touching the assign button 11f in Figure 7) (step S207).

[0084] The control unit 21 transmits the received approval information to the server 1 via the communication unit 23 (step S208). The approval information includes the task ID of the target task and the nurse ID of the nurse who assigned the target task. The control unit 11 of the server 1 receives the approval information transmitted from the nurse terminal 2 via the communication unit 13 (step S108).

[0085] The control unit 11 assigns the target task to a nurse according to the received approval information (step S109). The control unit 11 stores the nurse ID of the nurse to whom the target task was assigned in the task DB 154 of the large-capacity storage unit 15, associating it with the task ID of the target task (step S110).

[0086] The control unit 11 retransmits the task list, including the newly assigned target task, to the nurse terminal 2 via the communication unit 13 (step S111). The control unit 21 of the nurse terminal 2 receives the task list retransmitted from the server 1 via the communication unit 23 (step S209). The control unit 21 redisplays the received task list via the display unit 25 (step S210). The control unit 21 terminates processing.

[0087] <Example 1> This section describes the process of identifying which nurse to assign a task to, using the task assignment model 151, based on the assigned tasks currently held by multiple candidate nurses and the skill information of each nurse.

[0088] The task assignment model 151 in this modified example is a language model that uses nurse information, including the assigned tasks each nurse currently has and each nurse's skill information. The task assignment model 151 is trained with a large amount of training data. The training data includes the assigned tasks for each nurse and each nurse's skill information. Server 1 uses the training data to train the task assignment model 151. The training process is the same as in Embodiment 1, so the explanation is omitted.

[0089] Using the task assignment model 151 obtained through the learning process described above, the nurses to whom tasks are assigned are identified. Specifically, Server 1 retrieves the tasks already assigned to each nurse and the unassigned tasks from the task DB 154 based on the target date (for example, today).

[0090] Server 1 generates a prompt to be given to the task assignment model 151 based on the acquired assigned and unassigned tasks. The prompt includes the assigned tasks, the unassigned tasks, an instruction to identify the nurse to whom the tasks should be assigned, the nurse information for each nurse, and the output format of the identification results.

[0091] As an example, the generated prompt is: For the following unassigned tasks, refer to the tasks already assigned to each nurse and the information of multiple candidate nurses stored in Nurse DB152. Considering each nurse's areas of expertise, years of experience, skills, and work information (work days, work types, work hours), identify which nurse will be assigned to each task from among the multiple nurses. The identification results should be output in a format that shows the correspondence between the task and the assigned nurse. For example, the output format should be "Task ID: xxxx ⇒ Assigned Nurse: xxxx". Assigned tasks Nurse ID: Task ID N-001:T-0001 N-002:T-0002 N-002:T-0003 ×××××× Unassigned task list (to-do list): 1. Task ID: T-0010 Date: August 22 Time: 8:00 Task details: taking temperature Patient information: Patient ID:P-001 Medical department: Internal medicine Diagnosis: Diabetes 2. Task ID: T-0011 Date: August 22 Time: 8:00 Task details: taking temperature Patient information: Patient ID:P-002 Medical Department: Cardiology Diagnosis: Heart disease It could also be "xxxxxxx".

[0092] Server 1 inputs the generated prompt into the task assignment model 151 and outputs identification results that identify the nurse to whom each task will be assigned.

[0093] As an example, the specific result output was: Task ID: T-0010 ⇒ Nurse in Charge: N-003 Task ID: T-0011 ⇒ Nurse in Charge: N-004 It could also be "xxxxxxx".

[0094] Server 1 sends the specific result output from the task assignment model 151 to the nurse administrator's terminal 2. The administrator's nurse terminal 2 receives the specific result sent from Server 1. The administrator's nurse terminal 2 displays the received specific result on its screen. If the administrator approves each task, the administrator's nurse terminal 2 assigns each task to the appropriate nurse.

[0095] According to this modified version, it becomes possible to identify nurses to whom unassigned tasks will be assigned using a task assignment model 151 that uses nurse information, including tasks already assigned to each nurse and information on each nurse's skills.

[0096] <Modification 2> This section describes the process of identifying which nurse to assign a task to, using the task assignment model 151, based on nurse information including the working hours of multiple candidate nurses.

[0097] The duration of work refers to the working hours corresponding to the type of work, such as day shift or night shift. When nurses work long hours, they are more likely to experience physical fatigue. Therefore, it is necessary to assign tasks while taking into account the duration of nurses' work.

[0098] The task assignment model 151 in this modified example is a language model that uses tasks and nurse information, including the working hours of multiple nurses. The task assignment model 151 is trained with a large amount of training data. The training data includes tasks and nurse information, including the working hours of multiple nurses. Server 1 uses the training data to train the task assignment model 151. The training process is the same as in Embodiment 1, so the explanation is omitted.

[0099] Using the task assignment model 151 obtained through the learning process described above, the system identifies the nurse to whom the task will be assigned. Specifically, Server 1 retrieves tasks at the medical facility from the task DB 154 based on the target date (e.g., today). Server 1 calculates the work duration of several candidate nurses. For example, Server 1 calculates the work duration of each nurse based on their work start time (e.g., 8:00) and the current time.

[0100] Server 1 generates a prompt to be given to the task assignment model 151 based on the acquired task and the calculated work duration. The prompt includes the task, an instruction to identify the nurse to whom the task should be assigned from among several candidate nurses, nurse information including each nurse's work duration, and the output format of the identification result.

[0101] As an example, the generated prompt is: "For the following tasks, refer to the information of multiple candidate nurses stored in Nurse DB152, and consider each nurse's assigned subject, years of experience, skills, and work information (work days, work type, work hours, and work duration) to identify which nurse will be assigned to each task from among the multiple nurses." In particular, if your shift will last longer than 6 hours, please prioritize assigning tasks to other nurses. The identification results should be output in a format that shows the correspondence between the task and the assigned nurse. For example, the output format should be "Task ID: xxxx ⇒ Assigned Nurse: xxxx". Task list (To-Do list): 1. Task ID: T-0001 Date: August 22 Time: 8:00 Task details: taking temperature Patient information: Patient ID:P-001 Medical department: Internal medicine Diagnosis: Diabetes 2. Task ID: T-0002 Date: August 22 Time: 8:00 Task details: taking temperature Patient information: Patient ID:P-002 Medical Department: Cardiology Diagnosis: Heart disease It could also be "xxxxxxx".

[0102] Server 1 inputs the generated prompt into the task assignment model 151 and outputs identification results that identify the nurse to whom each task will be assigned.

[0103] As an example, the specific result output was: Task ID: T-0001 ⇒ Nurse in Charge: N-001 Task ID: T-0002 ⇒ Nurse in Charge: N-003 It could also be "xxxxxxx".

[0104] Server 1 sends the specific result output from the task assignment model 151 to the nurse administrator's terminal 2. The administrator's nurse terminal 2 receives the specific result sent from Server 1. The administrator's nurse terminal 2 displays the received specific result on its screen. If the administrator approves each task, the administrator's nurse terminal 2 assigns each task to the appropriate nurse.

[0105] According to this modified version, it becomes possible to identify the nurse to whom a task will be assigned using a task assignment model 151 that uses nurse information, including the working hours of each nurse.

[0106] <Variation 3> This section describes the process of identifying nurses to whom tasks should be assigned, using the task assignment model 151, based on the prioritization of multiple tasks. Tasks can be prioritized according to their urgency or importance, etc. Priorities include, for example, highest priority (high urgency), high priority, medium priority, and low priority.

[0107] The highest priority is for responding to emergencies requiring immediate attention (e.g., cardiac arrest, respiratory distress, or severe bleeding). High priority is for patients whose condition is rapidly deteriorating or who have severe symptoms (e.g., severe pain, high fever, or acute respiratory distress). Medium priority is for administering medication, performing procedures, or conducting tests. Low priority is for routine care (e.g., serving meals).

[0108] Figure 9 is an explanatory diagram showing an example of the record layout of Task DB154 in Modification Example 3. Note that the same reference numerals are used for elements that overlap with Figure 4, and their explanations are omitted. Task DB154 includes a priority column. The priority column stores the priority of the task. For example, the priority column stores "Highest (highest priority)", "Medium (medium priority)", and "Low (low priority)".

[0109] Server 1 retrieves tasks at the medical facility from Task DB 154 based on the target date (e.g., today). The tasks include task details, execution time, and priority. Based on the retrieved tasks, Server 1 generates prompts to be given to Task Assignment Model 151. The prompts include the task (task details, execution time, priority, etc.), an instruction to identify the nurse to whom the task should be assigned from among several candidate nurses, nurse information for each nurse, and the output format of the identification result.

[0110] As an example, the generated prompt is: Based on the priority of the following tasks, refer to the information of multiple candidate nurses stored in Nurse DB152, and consider each nurse's areas of expertise, years of experience, skills, and work information (work days, work types, work hours) to identify which nurse will be assigned to each task. The identification results should be output in a format that shows the correspondence between the task and the assigned nurse. For example, the output format should be "Task ID: xxxx ⇒ Assigned Nurse: xxxx". Task list (To-Do list): 1. Task ID: T-0001 Date: August 22 Time: 8:00 Task details: taking temperature Priority: Low Patient information: Patient ID:P-001 Medical department: Internal medicine Diagnosis: Diabetes 2. Task ID: T-0002 Date: August 22 Time: 8:00 Task details: taking temperature Priority: Low Patient information: Patient ID:P-002 Medical Department: Cardiology Diagnosis: Heart disease It could also be "xxxxxxx".

[0111] Server 1 inputs the generated prompt into the task assignment model 151 and outputs identification results that identify the nurse to whom each task will be assigned.

[0112] As an example, the specific result output was: Task ID: T-0001 ⇒ Nurse in Charge: N-001 Task ID: T-0002 ⇒ Nurse in Charge: N-002 It could also be "xxxxxxx".

[0113] Server 1 sends the specific result output from the task assignment model 151 to the nurse administrator's terminal 2. The administrator's nurse terminal 2 receives the specific result sent from Server 1. The administrator's nurse terminal 2 displays the received specific result on its screen. If the administrator approves each task, the administrator's nurse terminal 2 assigns each task to the appropriate nurse.

[0114] According to this modified version, by providing the task assignment model 151 with a prompt that includes the task, an instruction to identify the nurse to whom the task should be assigned from among several candidate nurses, nurse information of several nurses to be referenced, and the priority of several tasks, it becomes possible to identify the nurse to whom the task should be assigned.

[0115] Next, we will explain the retraining process for the task assignment model 151. In actual operation in a medical facility, the nurse identified by the task assignment model 151 may be unable to perform the task due to sudden absence, illness, or abrupt changes in duties. Therefore, flexible task management is necessary. Even if the task assignment model 151 identifies the optimal nurse, the identified nurse may not actually be able to perform the task, so it is necessary to assign the task to a different nurse than the one identified.

[0116] Returning to Figure 7, for example, if the nurse identified by the task assignment model 151 (for example, the nurse with nurse ID "N-002") is unable to perform the task, the nurse terminal 2 accepts the selection of a different nurse from the identified nurse by receiving a touch operation on the candidate nurse display field 11h within the identification result display field 11e by the administrator.

[0117] Nurse terminal 2 obtains the nurse ID of the selected nurse. For example, nurse terminal 2 may accept the selection of a nurse whose nurse ID is "N-001". When nurse terminal 2 receives a touch operation of the assign button 11f by the administrator, it sends approval information including the task ID and nurse ID to server 1. Subsequently, in the same manner as described above, server 1 assigns the target task to the selected nurse.

[0118] In this way, if the assigned nurse is changed to a different nurse than the one initially identified, Server 1 acquires retraining data that includes the changed nurse. The retraining data includes the nurse information for the task and multiple nurses, as well as the changed nurse to whom the task was assigned. The retraining data may also include the nurse information of the nurse before the change, the date and time the assignment was changed, and the reason for the change. Server 1 uses the acquired retraining data to retrain the task assignment model 151.

[0119] Specifically, Server 1 performs a conversion process to transform the acquired retraining data into a format (tokens) that the task assignment model 151 can process. That is, Server 1 performs training on a tokenizer that divides the acquired retraining data into tokens. Server 1 uses the divided tokens to retrain the task assignment model 151 for a specific task (e.g., question answering). Reinforcement learning may also be performed on the task assignment model 151.

[0120] According to this embodiment, it is possible to identify the nurse to whom a task will be assigned using the task assignment model 151.

[0121] According to this embodiment, it is possible to output information about the identified nurse to the nurse's administrator's nurse terminal 2.

[0122] According to this embodiment, it is possible to assign tasks to nurses identified by the task assignment model 151, provided that the administrator's approval is obtained.

[0123] According to this embodiment, it is possible to retrain the task assignment model 151 using retraining data that includes nurses other than the identified nurse.

[0124] (Embodiment 2) Embodiment 2 relates to a configuration in which a task assignment model 151 that does not use a language model is used to identify the nurse to whom a task will be assigned. Note that explanations of content that overlaps with Embodiment 1 will be omitted.

[0125] Figure 10 is an explanatory diagram illustrating the outline of the nurse identification process using the task assignment model 151. In this embodiment, the task assignment model 151 uses, for example, a DNN (Deep Neural Network(s)) with a convolutional layer to perform the process of identifying (estimating) the nurse to whom a task will be assigned. The task assignment model 151 has an input layer that accepts input of tasks at a medical facility and nurse information of multiple nurses, an output layer that outputs the result of identifying the nurse to whom the task will be assigned, and an intermediate layer that extracts feature quantities of the task and nurse information.

[0126] Tasks include the task details, such as examinations and emergency response cases instructed by a physician, and time information for performing each task. Nurse information includes at least one piece of information, such as nurse ID, workday, work type, work hours, and assigned specialty.

[0127] The input layer has multiple neurons that receive task and nurse information inputs and passes the input task and nurse information to the hidden layer. The hidden layer has multiple neurons that extract feature quantities from the task and nurse information and passes them to the output layer.

[0128] The hidden layer processes the task and nurse information input from the input layer through multiple hidden layers (e.g., fully connected layers), using activation functions such as the sigmoid function, tanh function, or ReLU (Rectified Linear Unit) function to connect the units of each layer and obtain effective features. The output layer includes, for example, a sigmoid function or a softmax function, and outputs a probability value of which nurse is assigned to the estimated task based on the features output from the hidden layers.

[0129] Server 1 performs training using training data. The training data consists of data on tasks at medical facilities, and combinations of nursing information for multiple nurses and the nurses assigned to those tasks. The training data is generated based on a large amount of task assignment information collected from medical facilities. Note that the training data may also be data created manually.

[0130] Server 1 inputs tasks at a medical facility and nursing information for multiple nurses into its input layer. After processing in the intermediate layer, it obtains an estimated probability value for the nurse from the output layer. The probability value is, for example, greater than 0 and less than 1.

[0131] Server 1 compares the probability value of nurses output from the output layer with the correct value in the training data, and optimizes the parameters used in the calculations in the hidden layer so that the output value from the output layer approaches the correct value. These parameters are, for example, the weights (connection coefficients) between neurons. The method of parameter optimization is not particularly limited, but for example, Server 1 uses backpropagation to optimize various parameters.

[0132] Server 1 performs the above processing using the training data to train the task assignment model 151. This makes it possible to build a model that can identify the nurse to whom a task is assigned. Alternatively, the above training process may be performed on another computer (not shown), and the trained task assignment model 151 may be used. Alternatively, instead of building the task assignment model 151, the nurse to whom a task is assigned may be identified by using a WEB API (Application Programming Interface) that uses a machine learning model.

[0133] When Server 1 acquires task and nurse information, it inputs the acquired task and nurse information into the task assignment model 151. Server 1 performs computational processing in the intermediate layer of the task assignment model 151 to extract features from the task and nurse information. Server 1 inputs the extracted features into the output layer of the task assignment model 151 and outputs a result that identifies the nurse to whom the task will be assigned.

[0134] As shown in the diagram, for a task (e.g., "assisting with toileting") and the nursing information of multiple nurses, the output identifies specific results where the probability values ​​for "Nurse A," "Nurse B," and "Nurse C" are "0.10," "0.82," and "0.08," respectively.

[0135] Alternatively, a predetermined threshold can be used to output a specific result. For example, if Server 1 determines that the probability value of "Nurse B" (0.82) is greater than or equal to a predetermined threshold (e.g., 0.80), it will output "Nurse B" as the specific result. Alternatively, instead of using the threshold described above, the task assignment model 151 may output the nurse corresponding to the highest probability value from the probability values ​​of each nurse estimated by the model as the specific result.

[0136] Furthermore, the task assignment model 151 is not limited to DNNs and may be implemented using other models such as logistic regression, SVM (Support Vector Machine), k-NN (k-Nearest Neighbor algorithm), decision trees, simple Bayesian classifiers, or random forests.

[0137] Figure 11 is an explanatory diagram illustrating the process of assigning tasks to nurses in Embodiment 2. Note that the same reference numerals are used for elements that overlap with Figure 7, and their explanations are omitted. This screen includes a candidate selection field 12a. The candidate selection field 12a is a field for receiving the selection of a nurse to whom a task will be assigned from among multiple candidate nurses.

[0138] Server 1 retrieves the tasks to be assigned to nurses from Task DB154 based on the target date (for example, the current day). Server 1 also retrieves the nurse information of multiple nurses who will be working from Nurse DB152 based on the target date. The nurse information includes at least one piece of information such as nurse ID, work day, work type, work hours, and assigned subject.

[0139] Server 1 inputs the acquired target task and the nursing information of multiple nurses into the task assignment model 151. Server 1 performs computational processing in the intermediate layer of the task assignment model 151 to extract feature quantities of the task and nursing information. Server 1 inputs the extracted feature quantities into the output layer of the task assignment model 151 and outputs an identification result that identifies the nurse to whom the target task will be assigned.

[0140] The identification results include the probability for each candidate nurse. However, in addition to probability, a score indicating the degree to which each candidate nurse is suitable for the task, or rank information indicating the priority of each candidate nurse, may also be used.

[0141] Server 1 sends the identification result to the nurse administrator's terminal 2. The administrator's nurse terminal 2 receives the identification result sent from Server 1. The administrator's nurse terminal 2 displays the received identification result in the candidate selection field 12a. As shown in the figure, the candidate selection field 12a displays "Nurse (N-002): 82% Nurse (N-001): 10% Nurse (N-003): 8%". Note that in Figure 11, multiple candidate nurses are displayed in order of decreasing probability, but this is not the only way. For example, multiple candidate nurses may be displayed in alphabetical order, etc.

[0142] When Nurse Terminal 2 receives a selection operation from the administrator in the candidate selection field 12a, it retrieves the selected nurse. When Nurse Terminal 2 receives a touch operation from the administrator on the assign button 11f, it sends approval information, including the task ID and nurse ID, to Server 1. Subsequently, similar to the process in Figure 7, Server 1 performs the approval process and resends the task list, including the newly assigned target task, to Nurse Terminal 2.

[0143] Figure 12 is a flowchart showing the processing procedure for assigning tasks to nurses in Embodiment 2. Note that elements that overlap with those in Figure 8 are denoted by the same reference numerals and their explanations are omitted.

[0144] After executing the process in step S103, the control unit 11 of server 1 retrieves the nurse information (nurse ID, work day, work type, work hours, or corresponding subject, etc.) of multiple nurses who will be on duty based on the target date (for example, today) from the nurse DB 152 of the large-capacity storage unit 15 (step S121). The control unit 11 inputs the target task and the nurse information of each nurse into the task assignment model 151 (step S122). The control unit 11 then executes the process in step S106.

[0145] According to this embodiment, it is possible to identify the nurse to whom a task will be assigned, based on the tasks at the medical facility and the output of a task assignment model 151 that uses the nurse information of multiple nurses (for example, the probability of each candidate nurse).

[0146] According to this embodiment, it becomes possible to apply the layout of a template file, in which access permissions are set in association with user identification information, to new tabular data.

[0147] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the invention is indicated by the claims, not in the sense described above, and all modifications within the sense and scope equivalent to the claims are intended.

[0148] The matters described in each embodiment can be combined with each other. Furthermore, the independent and dependent claims described in the claims can be combined with each other in any combination, regardless of the form of reference. In addition, the claims use a form in which claims referencing two or more other claims (multi-claim form), but are not limited to this. A form in which multi-claims referencing at least one multi-claim (multi-multi-claim) may also be used. [Explanation of Symbols]

[0149] 1. Information processing device (server) 11 Control Unit 12 Storage section 13 Communications Department 14 Reading section 15 Mass storage 151 Task Assignment Models 152 Nurse DB 153 Patient DB 154 Task DB 1a Portable storage medium 1b Semiconductor memory 1P Control Program 2. Information processing terminal (nurse terminal) 21 Control Unit 22 Memory section 23 Communications Department 24 Input section 25 Display section 2P control program B Bus N Network

Claims

1. Obtain tasks at a medical facility, Based on the acquired tasks and the output from a model that uses healthcare professional information from multiple healthcare professionals, the healthcare professionals to whom the tasks will be assigned will be identified. A program that instructs a computer to perform a process.

2. The aforementioned model uses healthcare professional information, including the assigned tasks that each healthcare professional currently has, and the skills information of each healthcare professional. The program according to claim 1.

3. The aforementioned model further uses healthcare professional information, including the working hours of each healthcare professional. The program according to claim 1 or 2.

4. The aforementioned tasks include the task content, which includes examinations and emergency response cases instructed by a physician, and time information for performing each task. Based on the output from a model that uses the task content and time information, along with the healthcare professional information of multiple healthcare professionals, the healthcare professionals to whom the task will be assigned are identified. The program according to claim 1 or 2.

5. The information regarding the identified healthcare worker is output to the terminal device of the healthcare worker's administrator. The task will be assigned to the medical professional upon approval from the aforementioned administrator. The program according to claim 1 or 2.

6. We accept selection of unassigned tasks. The identified healthcare professionals and their assigned tasks are displayed. Approval for the medical professional in question will be accepted. The program according to claim 1 or 2.

7. The model is trained using training data that includes the task, the healthcare professional information of the multiple healthcare professionals, and the healthcare professionals to whom the task is assigned. The program according to claim 1 or 2.

8. The aforementioned model is a language model, The language model is provided with the acquired task, an instruction to identify the healthcare professional to assign the task from among several candidate healthcare professionals, and a prompt containing healthcare professional information of several healthcare professionals to be referenced, thereby identifying the healthcare professional to assign the task. The program according to claim 1 or 2.

9. We obtained retraining data that included healthcare professionals different from the identified healthcare professionals. The model is retrained using the acquired retraining data. The program according to claim 1 or 2.

10. The aforementioned model is a language model, The language model is provided with a prompt that includes the acquired task, an instruction to identify the healthcare professional to assign the task from among several candidate healthcare professionals, healthcare professional information of several healthcare professionals to be referenced, and the priority of several tasks, thereby identifying the healthcare professional to assign the task. The program according to claim 1 or 2.

11. Obtain tasks at a medical facility, Based on the acquired tasks and the output from a model that uses healthcare professional information from multiple healthcare professionals, the healthcare professionals to whom the tasks will be assigned will be identified. Information processing methods.

12. An information processing device comprising a control unit, The control unit, Obtain tasks at a medical facility, Based on the acquired tasks and the output from a model that uses healthcare professional information from multiple healthcare professionals, the healthcare professionals to whom the tasks will be assigned will be identified. Information processing device.

Images