Program, information processing method, information processing device and model generation method

Abstract

To provide a program which can present a recovery treatment to do on a patient who has become sicker after taking an administered medicine.SOLUTION: A computer acquires the vital data of a patient who has become in bad shape after an injection was administered or being administered to the patient and the type of the injection and the amount and the rate of the dosage according to a program. When the type of the injection, the amount and the rate of the dosage, the vital data are input, the computer inputs the acquired type of the injection and the acquired amount and rate of the dosage and the acquired vital data into a learning model which has learned to output advice information on a recovery treatment and outputs the advice information on the recovery treatment on the patient with the vital data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a program, an information processing method, an information processing device, and a model generation method. [Background technology]

[0002] Systems have been proposed that assist doctors in diagnosing patients' conditions at medical institutions, etc. For example, Patent Document 1 discloses a system that estimates a possible diagnosis or a patient's condition based on medical information including the patient's medical examination records, nursing records, test results, medication records, medical images, etc. The system disclosed in Patent Document 1 estimates and presents a patient's diagnosis based on the contents of a drug prescription and injection order made for the patient, and also outputs a warning if an ordered drug is contraindicated for the condition of the diagnosis based on the diagnosis and the contents of the drug prescription and injection order. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2007-18460 Summary of the Invention [Problem to be solved by the invention]

[0004] Medical institutions and other facilities handle a wide variety of medications, each with different dosages and administration rates. Therefore, when administering medications to patients, meticulous care is required to avoid mistaking medications or setting the wrong dosage or administration rate. Because an incorrect medication type, dosage, or administration rate can lead to a patient's serious condition, medical professionals are required to perform accurate work, placing a heavy burden on them. Furthermore, if an incorrect medication type, dosage, or administration rate is administered and the patient's condition subsequently worsens, recovery measures must be taken. However, medical professionals such as doctors and nurses have limited experience and knowledge, making appropriate decisions can be difficult. The system disclosed in Patent Document 1 associates a diagnosis or patient condition with the names of available or contraindicated medications, thereby inferring a diagnosis from the medication ordered for the patient and determining whether the ordered medication is a contraindicated medication. However, it cannot recommend appropriate measures for patients whose condition worsens after administration.

[0005] In one aspect, an object of the present invention is to provide a program or the like that can suggest recovery measures to be taken by a patient whose condition has worsened after medication. [Means for solving the problem]

[0006] A program according to one aspect acquires vital data of a patient who has become abnormal during or after administration of an injectable agent, as well as the type, dosage, and administration rate of the injectable agent; the acquired type, dosage, and administration rate of the injectable agent and the vital data are input into a learning model that has been trained to output advice information regarding recovery measures when the type, dosage, and administration rate of the injectable agent and the vital data are input; and the program causes a computer to execute a process of outputting advice information regarding recovery measures for the patient based on the vital data. [Effects of the Invention]

[0007] In one aspect, a patient whose condition worsens after taking a medication can be suggested to undergo remedial action. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a block diagram illustrating an example of the configuration of an information processing device. [Figure 2] FIG. 2 is a schematic diagram showing an example of the configuration of an electronic medical record DB. [Figure 3] FIG. 1 is a schematic diagram illustrating an example of the configuration of a learning model. [Figure 4] 10 is a flowchart illustrating an example of a learning model generation processing procedure. [Figure 5] 10 is a flowchart illustrating an example of a recovery measure presentation process; [Figure 6] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 7] FIG. 10 is a schematic diagram showing an example of the configuration of a learning model according to the second embodiment. [Figure 8] 10 is a flowchart showing an example of a recovery measure presentation processing procedure according to the second embodiment. [Figure 9] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 10] 13 is a flowchart showing an example of a recovery measure presentation processing procedure according to the third embodiment. [Figure 11] FIG. 10 is an explanatory diagram showing an example of a screen. [Figure 12] FIG. 1 is an explanatory diagram showing an example of the configuration of a learning model used to determine the state of a patient. [Figure 13] 13 is a flowchart showing an example of a recovery measure presentation processing procedure according to the fourth embodiment. [Figure 14] FIG. 10 is an explanatory diagram showing an example of a screen. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, a program, an information processing method, an information processing device, and a model generation method according to the present disclosure will be described in detail with reference to the drawings illustrating embodiments thereof.

[0010] (Embodiment 1) This description describes an information processing device that provides advice on recovery measures (restorative measures) for patients whose vital data becomes abnormal during or after the administration of an injectable drug, based on the vital data and medication information including the type (name) of the administered injectable drug, the dosage, and the administration rate. In this embodiment, the device provides restorative measures to be taken for patients whose condition worsens during or after the administration of an injectable drug used in a medication method in which a drug is administered directly into a blood vessel using, for example, an infusion pump or a syringe pump. Note that the injectable drug is not limited to drugs administered directly into a blood vessel, but may also be drugs administered intradermally, subcutaneously, or into body tissue, such as muscle, using a syringe needle. A worsening condition refers to a state in which, for example, blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, or body temperature included in the vital data deviates from the normal range. Note that the normal range is preset based on, for example, age, gender, weight, height, medical condition, etc.

[0011] FIG. 1 is a block diagram showing an example of the configuration of an information processing device. The information processing device 10 is a device capable of various information processing and transmitting and receiving information, such as a server computer or a personal computer. The information processing device 10 may be configured to have multiple devices for distributed processing, or may be realized by multiple virtual machines installed in a single device. The information processing device 10 is installed and used, for example, in a medical institution. When the information processing device 10 is configured as a server computer, the information processing device 10 may be a local server installed in the medical institution, or may be a cloud server connected for communication via a network such as the Internet.

[0012] The information processing device 10 performs a process of outputting advice information that advises on recovery measures to recover the patient's condition after medication, based on medication information regarding a drug (injection) being administered to the patient or after administration and vital data of the patient before and after medication. Specifically, as described below, the information processing device 10 performs machine learning to learn predetermined training data, and prepares in advance a learning model 12M (see FIG. 3) that inputs medication information and vital data and outputs advice information. The information processing device 10 inputs the medication information and vital data of a patient whose condition has worsened during or after medication into the learning model 12M, and obtains advice information from the learning model 12M that suggests measures to recover the condition.

[0013] The information processing device 10 includes a control unit 11, a storage unit 12, a communication unit 13, an input unit 14, a display unit 15, a reading unit 16, etc., and these units are connected to each other via a bus. The control unit 11 has one or more processors such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), a GPU (Graphics Processing Unit), or an AI chip (AI semiconductor). The control unit 11 appropriately executes a control program 12P stored in the storage unit 12, thereby performing various information processing, control processing, etc. that the information processing device 10 should perform.

[0014] The storage unit 12 includes a RAM (Random Access Memory), a flash memory, a hard disk, an SSD (Solid State Drive), etc. The storage unit 12 pre-stores a control program 12P (program product) executed by the control unit 11 and various data necessary for executing the control program 12P. The storage unit 12 also temporarily stores data generated when the control unit 11 executes the control program 12P. The storage unit 12 also stores a learning model 12M that has learned training data through, for example, machine learning. The learning model 12M is a trained model that has been trained to output advice information, including recovery measures for recovering a patient's condition after medication, when the patient's medication information and vital data are input. The learning model 12M is expected to be used as a program module constituting artificial intelligence software. The learning model 12M performs a predetermined calculation on input values ​​and outputs the calculation results. The storage unit 12 stores data, such as coefficients and thresholds of functions that define this calculation, as the learning model 12M. The storage unit 12 also stores an electronic medical record DB 12a, which will be described later. The electronic medical record DB 12a may be stored in another storage device connected to the information processing device 10, or may be stored in another storage device with which the information processing device 10 can communicate.

[0015] The communication unit 13 is a communication module for connecting to a network such as the Internet or a LAN (Local Area Network) by wired or wireless communication, and transmits and receives information to other devices via the network. The input unit 14 accepts operation input by a user and sends a control signal corresponding to the operation content to the control unit 11. The display unit 15 is a liquid crystal display, an organic EL display, or the like, and displays various information according to instructions from the control unit 11. The input unit 14 and the display unit 15 may be a touch panel configured as an integrated unit.

[0016] The reading unit 16 reads information stored in a portable storage medium 1a, such as a CD (Compact Disc)-ROM, a DVD (Digital Versatile Disc)-ROM, a USB (Universal Serial Bus) memory, or an SD (Secure Digital) card. The control program 12P (program product) and various data stored in the storage unit 12 may be read by the control unit 11 from the portable storage medium 1a via the reading unit 16 and stored in the storage unit 12. The control program 12P and various data stored in the storage unit 12 may also be downloaded by the control unit 11 from another device via the communication unit 13 and stored in the storage unit 12.

[0017] FIG. 2 is a schematic diagram showing an example of the configuration of the electronic medical record DB 12a. The electronic medical record DB 12a is a database that stores electronic medical record data (medical records) of patients who use medical institutions. The electronic medical record DB 12a shown in FIG. 2 includes a patient ID column, a patient information column, a vital data column, a medication information column, and the like. The patient ID column stores identification information (patient ID, patient code) for identifying each patient. The patient ID may be, for example, the patient card number of a patient card issued by a medical institution. The patient information column stores patient information related to a patient in association with the patient ID. The patient information includes, for example, attribute information including the patient's age and gender, diagnosis, medical history, height, weight, and the like. The vital data column stores vital data measured from a patient in association with the patient ID. The vital data includes, for example, blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, body temperature, and the like, and these data are stored in each column in association with, for example, the measurement date and time. Each piece of vital data may be represented by a graph showing changes over time, and the graph data is a series of data, for example, pairs of measurement date and time and measured values, recorded each time a measurement is taken. The medication information sequence stores medication information relating to injections (drugs) that have been administered to a patient or are scheduled to be administered, in association with a patient ID. The medication information includes, for example, the date and time of administration, the name of the drug (type of drug), the dosage, the administration rate, the progress during or after administration, and effective recovery measures taken if the progress worsens. The progress status includes a normal state (good) and an abnormal state. An abnormal state is, for example, a state in which vital data (blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, body temperature, etc.) deviates from the normal range after the start of medication, including, for example, blood pressure rising above the normal range or falling below the normal range, heart rate rising above the normal range or falling below the normal range, respiratory rate rising above the normal range or falling below the normal range, abnormal electrocardiogram waveforms, body temperature rising above the normal range or falling below the normal range, etc. Recovery measures include, for example, stopping (interrupting) the medication in progress, increasing or decreasing the dosage of the medication in progress, increasing or slowing the administration rate of the medication in progress, administering an antagonist to the medication in progress, administering a different medication, etc.

[0018] The contents stored in the electronic medical record DB 12a are not limited to the example shown in Fig. 2, and various types of information related to the patient may be stored, such as examination information related to various examinations such as blood tests and urine tests conducted on the patient, treatment information related to treatment, surgery information related to surgery, medication history, comments entered by doctors, etc. The examination information may include medical images such as X-ray images, ultrasound images, CT (Computed Tomography) images, MRI (Magnetic Resonance Imaging) images, and PET (Positron Emission Tomography) images.

[0019] FIG. 3 is a schematic diagram showing an example of the configuration of the learning model 12M. The learning model 12M shown in FIG. 3 is trained to input a patient's medication information (drug name, dosage, and administration rate) and vital signs (blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, etc.), perform a calculation to determine the recovery procedure to be performed on the patient based on the input data, and output the calculation result (advice information indicating the recovery procedure). The learning model 12M may be configured to input the patient's medication information stored in the electronic medical record DB 12a and data included in the electronic medical record data. The learning model 12M is configured, for example, by a convolutional neural network (CNN), which is a neural network model generated by deep learning. The learning model 12M may be configured using algorithms other than CNN, such as a recurrent neural network (RNN), a long short-term memory (LSTM), a transformer, a decision tree, a random forest, or a support vector machine (SVM), or may be configured by combining multiple algorithms. The information processing device 10 performs machine learning to learn predetermined training data to generate a learning model 12M in advance. For a patient whose condition worsens during or after medication, the information processing device 10 inputs the patient's medication information and vital data before and after medication registered in the electronic medical record DB 12a into the learning model 12M, and predicts the recovery procedure to be performed on the patient based on the output information from the learning model 12M. The learning model 12M has multiple input nodes, each associated with information to be input. The medication information and vital data are input to the learning model 12M via the associated input nodes.

[0020] The learning model 12M shown in FIG. 3 has multiple output nodes, each associated with a preset recovery action. Each output node outputs a probability that the associated recovery action should be determined to be the action to be performed. Recovery actions include stopping the administration of the drug being administered, increasing the dosage of the drug being administered, increasing or decreasing the administration rate of the drug being administered, administering an antagonist to the drug being administered (antagonist A1, administered at dosage B1 and administration rate C1), administering a different drug (drug A3, administered at dosage B3), etc. In the example shown in FIG. 3, output node 0 outputs a probability that the administration should be discontinued, output node 1 outputs a probability that the dosage should be increased, and output node 2 outputs a probability that the administration rate should be increased. Note that the recovery actions associated with each output node are not limited to the example shown in FIG. 3 and include recovery actions to be performed when any vital signs deviate from the normal range. The output value of each output node is, for example, a value between 0 and 1.0, and the sum of the discrimination probabilities output from each output node is 1.0 (100%). With the above-mentioned configuration, when a patient's medication information and vital data are input, the learning model 12M of this embodiment outputs information related to recovery measures (discrimination probability for each recovery measure) for recovering from the condition (worsened condition) indicated by the input vital data.

[0021] In the learning model 12M described above, the information processing device 10 selects, for example, a predetermined number (e.g., three) of output values ​​(discrimination probabilities) from among the output values ​​from each output node in descending order of magnitude, and specifies the recovery action associated with the output node that output the selected output value as the action to be performed. This makes it possible to present a predetermined number of recovery actions as candidates for the recovery action to be performed. Note that the information processing device 10 may also specify, as the action to be performed, the recovery action associated with the output node that output the largest output value (discrimination probability) among the output values ​​from each output node of the learning model 12M. Furthermore, the learning model 12M may be configured to have a single output node that outputs the recovery action with the highest discrimination probability, instead of having multiple output nodes that output discrimination probabilities for each recovery action.

[0022] The learning model 12M is generated by machine learning an untrained learning model using training data including training medication information and vital data, and information indicating effective recovery measures (correct answer labels). The training data is generated by assigning information indicating effective recovery measures to medication information and vital data before and after administration for patients who have been administered various injectable drugs and whose condition worsened during or after administration. The medication information, vital data, and effective recovery measures can be acquired from the medication information, vital data, and recovery measures of patients who have already administered medication, stored in the electronic medical record DB 12a. Specifically, the medication information, vital data before and after administration, and recovery measures performed for patients whose condition worsened and whose vital data returned to a normal range after recovery measures were implemented are read from the medication information stored in the electronic medical record DB 12a to generate the training data.

[0023] When training medication information and vital data are input, the learning model 12M learns so that the output value from the output node corresponding to the recovery action indicated by the correct label approaches 1.0 and the output values ​​from other output nodes approach 0.0. In the learning process, the learning model 12M performs calculations based on the input medication information and vital data to calculate output values ​​from each output node. The learning model 12M then compares the calculated output value of each output node with a value corresponding to the correct label (1 for the output node corresponding to the recovery action indicated by the correct label, and 0 for other output nodes), and optimizes parameters used in the calculation process so that each output value approximates the value corresponding to the correct label. The parameters are, for example, weights between neurons in the learning model 12M. The parameter optimization method is not particularly limited, and examples include backpropagation and steepest descent. As a result, a learning model 12M is obtained that is trained to predict the recovery action to be performed on a patient and output the prediction result when a patient's medication information and vital data are input.

[0024] The learning model 12M may be trained by another learning device. The trained learning model 12M generated by training on another learning device is downloaded from the learning device to the information processing device 10 via a network or a portable storage medium 1a, for example, and stored in the storage unit 12. The learning model 12M is not limited to the configuration shown in FIG. 3. The learning model 12M may be configured to receive various information used in predicting the recovery procedure to be performed. For example, the learning model 12M may receive one or more pieces of information, such as height and weight, medical history, examination information, treatment information, surgery information, medication history, and comments entered by a doctor or the like, contained in the electronic medical record data. In this case, the learning model 12M is configured to determine the recovery procedure to be performed based on not only the medication information and vital sign data but also various pieces of information contained in the electronic medical record data, and to output a discrimination probability for each recovery procedure. The electronic medical record data input to the learning model 12M may be time-series data, such as time-series data (examination information) of test results of blood tests or urine tests that are periodically performed, or time-series data (treatment information) of various treatments that are periodically performed. In this case, for example, the electronic medical record data of the patient for the most recent day, or the electronic medical record data for the most recent week, etc. may be configured to be input to the learning model 12M.

[0025] The process of learning the training data and generating the learning model 12M will be described below. Fig. 4 is a flowchart showing an example of the process procedure for generating the learning model 12M. The following process is performed by the control unit 11 of the information processing device 10 in accordance with the control program 12P stored in the storage unit 12, but may also be performed by another learning device.

[0026] The control unit 11 of the information processing device 10 first acquires patient information to be used for training data from the electronic medical record DB 12a, generates training data, and uses the generated training data to train the learning model 12M. For example, the control unit 11 determines whether any patient's condition has worsened after the start of medication (during or after medication) based on the progress of medication information for each patient stored in the electronic medical record DB 12a (S11). Specifically, the control unit 11 determines whether any patient's vital data (blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, body temperature, etc.) has deviated from the normal range within a predetermined time (e.g., several tens of minutes, several hours, etc.) since the start of medication. If the control unit 11 determines that any patient's condition has worsened after the start of medication (S11: YES), the control unit 11 acquires the patient's medication information, vital data before and after medication, and information indicating recovery measures taken after the patient's condition has worsened from the electronic medical record DB 12a (S12). The control unit 11 acquires vital data before and after medication, for example, from a predetermined time before the start of medication (several minutes or several tens of minutes) until the patient's condition worsens and then recovers. The control unit 11 generates training data by assigning a correct answer label indicating the acquired recovery measure to the acquired medication information and vital data, and stores the training data in the storage unit 12 (S13).

[0027] The control unit 11 determines whether there are any patients (unprocessed patients) for whom the training data generation process has not been performed among the patients whose condition worsened after the start of medication (S14). If it is determined that there are any unprocessed patients (S14: YES), the control unit 11 returns to the process of step S12 and acquires the medication information, vital data before and after medication, and information indicating recovery measures of the unprocessed patients from the electronic medical record DB 12a (S12), and stores these as training data in the storage unit 12 (S13). The control unit 11 repeats the processes of steps S12 to S14 until it determines that there are no unprocessed patients. As a result, training data to be used for learning the learning model 12M can be generated and accumulated based on the information of patients whose condition worsened after the start of medication.

[0028] If it is determined that there is no patient whose condition worsened after the start of medication (S11: NO), the control unit 11 ends the process without performing the above-mentioned process. Note that, for patients whose condition worsened after the start of medication, medication information, vital data before and after medication, and information on the recovery measures taken may be collected in advance, and the control unit 11 may generate training data to be used for learning the learning model 12M based on the collected data.

[0029] If it is determined that there are no untreated patients (S14: NO), the control unit 11 performs a learning process for the learning model 12M using each of the accumulated training data (S15). Here, the control unit 11 inputs the medication information and vital data contained in the training data into the learning model 12M and acquires an output value from each output node. The control unit 11 compares the output value of each output node with a value corresponding to the correct label (1 for the output node corresponding to the correct recovery action, and 0 for the output node corresponding to the incorrect recovery action), and optimizes parameters such as the weights between neurons in the learning model 12M using, for example, backpropagation so that the two values ​​approximate each other.

[0030] The control unit 11 determines whether or not there is any training data (unprocessed data) that has not yet undergone learning processing among the training data accumulated in step S13 (S16). If it is determined that there is unprocessed data (S16: YES), the control unit 11 returns to the processing of step S15 and repeats the learning processing using the unprocessed training data. If it is determined that there is no unprocessed data (S16: NO), the control unit 11 ends the series of processing.

[0031] The above-described process generates a learning model 12M trained to output information indicating the recovery procedure to be performed on a patient when the patient's medication information and vital data are input. The learning model 12M can be further optimized by repeatedly performing the learning process using the training data described above. An already trained learning model 12M can also be retrained by performing the above-described process, in which case a learning model 12M with higher discrimination accuracy can be generated. Alternatively, a learning model 12M tailored to each medical institution may be generated by training data for each medical institution, for example. The recovery procedures that can be performed on a patient vary somewhat depending on the medical instruments or drugs available at each medical institution. Therefore, generating a learning model 12M for each medical institution makes it possible to determine the recovery procedure to be performed on the patient, taking into account the recovery procedures that are frequently performed at each medical institution.

[0032] The following describes a process for determining recovery measures to be taken on a patient and outputting advice when, for example, the condition of a patient taking medication worsens, using the learning model 12M generated by the above-described process. FIG. 5 is a flowchart showing an example of the recovery measures presentation process, and FIG. 6 is an explanatory diagram showing an example screen. The following process is performed by the control unit 11 of the information processing device 10 in accordance with the control program 12P stored in the memory unit 12.

[0033] If a patient's condition worsens during or after administration of a drug in a medical setting, a doctor or nurse can receive advice on recovery measures to be taken by entering medication information (drug name, dosage, and administration rate) regarding the medication being administered or after administration via an input screen such as that shown in FIG. 6A. Medication is administered to patients in, for example, an operating room, an ICU (intensive care unit), a hospital room, or a treatment room. For simplicity's sake, the following description will be given assuming that the information processing device 10 is installed in a room where medication is administered. Alternatively, a terminal having an input unit and a display unit may be installed in the room where medication is administered, and this terminal may be communicatively connected to the information processing device 10 in the medical institution. In this case, medication information input via the terminal is transmitted to the information processing device 10, and recovery measures predicted by the information processing device 10 based on the medication information are transmitted to the terminal and presented to the doctor or nurse via the display unit of the terminal.

[0034] The control unit 11 of the information processing device 10 displays an input screen as shown in FIG. 6A on the display unit 15 in accordance with, for example, an operation by a doctor or nurse via the input unit 14 (S21). The screen shown in FIG. 6A has an input field for a patient ID, input fields for medication information such as a drug name, dosage, and administration rate, and an OK button for specifying a recovery procedure based on the input medication information. Each input field may be configured to allow input of any information or numerical value, or may have a pull-down menu for selecting any one of multiple options. In addition to the configuration shown in FIG. 6A, the input screen may also have an input field for inputting a drug manufacturer, and the input field for drug name may be configured to, for example, display a list of drug names beginning with the input letter or number when any letter or number is input.

[0035] The control unit 11 acquires the patient ID input via, for example, the input unit 14 and displays it in an input field for the patient ID (S22). Note that, for example, if a code reader such as a barcode reader or a QR Code (registered trademark) reader is connected to the information processing device 10, the control unit 11 may read the patient ID using the code reader and display it in the input field. For example, if the patient is wearing a wristband on which a code encoding the patient ID is printed, the patient ID may be read from the code on the wristband using the code reader, or the patient ID may be read from a sheet of paper on which the patient ID code is printed.

[0036] The control unit 11 (acquisition unit) also acquires medication information input via, for example, the input unit 14 and displays it in each input field (S23). The control unit 11 may read medication information related to the currently administered medication or the most recently administered medication from the patient's electronic medical record data stored in the electronic medical record DB 12a based on the input patient ID, and display it in each input field. When administering medication using an infusion pump or a syringe pump, the drug name, dosage, and administration rate are set in the pump. Therefore, if the information processing device 10 is configured to be able to communicate with the pump, the control unit 11 may acquire the drug name, dosage, and administration rate set in the pump from the pump by transmitting and receiving information to and from the pump, and display them in the input fields. Acquiring each piece of information from the pump prevents input errors and reduces the operational burden on doctors and nurses. Furthermore, if a sticker bearing a code printed with coded drug information, including the drug name, is affixed to the drug package, the control unit 11 may read the drug name from the code on the drug package using a code reader. In addition, if an RFID (Radio Frequency IDentifier) ​​tag storing drug information including the drug name is attached to the drug, and a reading device that reads the stored information in the RFID tag is connected to the information processing device 10, the control unit 11 may read the drug name from the RFID tag using the reading device.

[0037] The control unit 11 determines whether the OK button on the input screen has been operated (S24). If it determines that the OK button has not been operated (S24: NO), the process returns to step S23 and repeats the acquisition and display of medication information. If it determines that the OK button has been operated (S24: YES), the control unit 11 (acquisition unit) acquires vital data of the patient from the electronic medical record DB 12a (S25). Here, based on the patient ID (patient identification code) entered on the input screen, the control unit 11 acquires vital data measured at a predetermined time most recently from vital data stored in association with the patient ID. Note that the control unit 11 acquires vital data measured, for example, from a predetermined time (several minutes to several tens of minutes) before the start of medication to immediately before. The vital data includes blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, etc.

[0038] The control unit 11 inputs the medication information entered via the input screen and the vital data acquired in step S25 into the learning model 12M, and identifies (predicts) recovery measures to be implemented on the patient based on the output values ​​from the learning model 12M (S26). For example, the control unit 11 selects a predetermined number (e.g., three) of output values ​​(discrimination probabilities) from the output values ​​of each output node in the learning model 12M in descending order of magnitude, and identifies the recovery measures associated with the output node that output the selected output values ​​as candidates for recovery measures to be implemented. Then, the control unit 11 generates advice information that presents the identified candidate recovery measures to a doctor or nurse (S27). For example, the control unit 11 generates advice information that lists messages explaining the recovery measures identified in step S26. Specifically, the control unit 11 generates advice information including messages such as "Stop medication," "Increase the dosage of the drug being administered," "Increase (or decrease) the administration rate of the drug being administered," "Administer antagonist A1 at dosage B1 and administration rate C1," "Administer drug A3 at dosage B3," and "Watch and see how it goes."

[0039] Then, the control unit 11 (output unit) displays an advice screen as shown in FIG. 6B on the display unit 15 based on the generated advice information (S28). The advice screen shown in FIG. 6B presents three recovery measures, in descending order of discriminant probability, as candidate recovery measures to be implemented on the patient. The control unit 11 may generate advice information suggesting the implementation of multiple recovery measures based on the discriminant probability for each recovery measure. For example, if the highest discriminant probability and the second highest discriminant probability are similar in value, the control unit 11 may generate advice information suggesting the implementation of these two recovery measures. In the example shown in FIG. 6B, recommendation 2 proposes two recovery measures: "discontinue administration" and "administer antagonist A1 at dose B1 and at administration rate C1." The advice screen may also display the discriminant probability output from the output node corresponding to each recovery measure, in association with each recovery measure. In this case, the degree of recommendation for each recovery measure can be presented as a discriminant probability. The physician confirms the recommended recovery measures on the advice screen, selects an appropriate recovery measure, and implements the recovery measure on the patient. If the information processing device 10 has an audio output unit such as a speaker, the control unit 11 may present the advice information to the doctor or nurse by outputting the advice information as audio in addition to displaying it. Furthermore, if a doctor is present in a separate room, etc., by providing another terminal capable of communicating with the information processing device 10 in the separate room, the nurse who administered the medication can receive instructions to perform recovery processing from the doctor via the terminal provided in the separate room while viewing the advice screen using the information processing device 10. Specifically, the information processing device 10 transmits the advice screen to the terminal in the separate room, and the doctor in the separate room can determine an appropriate recovery processing from the recovery processing presented on the advice screen and instruct the nurse to perform the determined recovery processing via the terminal in the separate room.

[0040] The above-described process allows for the provision of advice on recovery measures to be taken by a patient to improve their condition if their condition worsens during drug administration. Therefore, doctors can quickly determine the appropriate recovery measures by considering the presented recovery measures. If a patient's condition worsens due to medication, it becomes an emergency situation with a high degree of urgency, making it difficult to make an appropriate decision. Furthermore, when a limited number of doctors are available, such as at night, it is possible that doctors with more experience and knowledge may not be able to provide appropriate support. Even in such situations, the information processing device 10 of this embodiment allows doctors to appropriately determine the recovery measures to be taken by the patient, and nurses to implement appropriate recovery measures according to the doctor's instructions, thereby enabling the patient's condition to recover quickly. Therefore, the burden on medical professionals, such as doctors and nurses, who must determine appropriate recovery measures in urgent situations can be reduced. Furthermore, in this embodiment, the learning model 12M is used to identify the recovery measures to be taken by a patient based on the patient's medication information and vital signs. Therefore, when using a learning model 12M trained on a large amount of training data, more appropriate recovery measures can be identified.

[0041] In this embodiment, since an appropriate recovery procedure can be determined early, the patient's deteriorating condition can be quickly alleviated, and the impact on the patient can be minimized. By minimizing the impact on the patient, further recovery procedures become unnecessary, and the use of medical resources, including medical equipment, drugs, and human resources (medical professionals such as doctors and nurses), used for the unnecessary recovery procedures can be reduced.

[0042] In the present embodiment, the information processing device 10 is described as locally performing a process for identifying a recovery procedure to recover a patient's worsening condition using the learning model 12M. However, this configuration is not limited to this. For example, the information processing device 10 may be configured as a server computer, and the process for identifying a recovery procedure using the learning model 12M may be performed by the server computer. In this case, the server computer may be configured to acquire the patient's medication information and vital data from a terminal device used by a doctor or nurse connected via a network, and output and display information on the identified recovery procedure to the terminal device. Note that the server computer may store the electronic medical record DB 12a. In this case, the server computer only needs to acquire the patient's medication information from the terminal device. Alternatively, a first server that performs the process for identifying a recovery procedure using the learning model 12M and a second server that stores the electronic medical record DB 12a may be separately provided. In this case, the first server may be configured to acquire the patient's medication information and vital data from the second server and perform the process for identifying a recovery procedure using the learning model 12M. Even with this configuration, the same processing as in the above-described embodiment is possible, and the same effects can be obtained.

[0043] (Embodiment 2) This section describes an information processing device that, when predicting a recovery procedure for recovery from a patient's medication information and vital data, identifies input data that has a large contribution to the obtained prediction result (discrimination result) from among the medication information and vital data input to a learning model 12M, and presents the input data together with the prediction result. The information processing device of this embodiment has a configuration similar to that of the information processing device 10 of embodiment 1, so a detailed description of the configuration will be omitted. Note that, in the information processing device 10 of this embodiment, the learning model stored in the storage unit 12 is slightly different from the learning model 12M of embodiment 1 shown in FIG. 3.

[0044] FIG. 7 is a schematic diagram showing an example of the configuration of a learning model according to the second embodiment. The learning model 12Ma according to this embodiment is a so-called explainable AI (XAI), a machine learning model that receives a patient's medication information and vital signs, and outputs information on recovery measures to be implemented on the patient (the discrimination probability of each recovery measure) and explanatory data indicating the basis for the prediction. In this embodiment, the learning model 12Ma may also be configured to receive medication information and electronic medical record data. The learning model 12Ma is configured to calculate, for each input data, a contribution (SHAP value, Attention, etc.) that serves as the basis for obtaining a prediction result (model output) using a SHAP or Attention mechanism, etc. The contribution is a value indicating the degree to which each input data of the learning model 12Ma contributed to the output of the predicted recovery measure (prediction result). For example, Attention is a value indicating the degree of attention paid to each input data (input element) in order to predict the prediction result. Therefore, input data with a high contribution can be considered the basis for obtaining the prediction result. The learning model 12Ma shown in Fig. 7 has the same configuration as the learning model 12M shown in Fig. 3, and in addition has a plurality of output nodes (the number of input data) corresponding to each input data as output nodes for explanation data. Therefore, the learning model 12Ma is configured to output contributions (explanation data) for each input data from each output node for explanation data. For example, output node 101 outputs the contribution for input data 1 (e.g., input data for input node 0), and output node 102 outputs the contribution for input data 2 (e.g., input data for input node 1).

[0045] The information processing device 10 selects a predetermined number (e.g., three) of output values ​​(contribution degrees) from each output node for explanation data in descending order, and identifies the input data associated with the output node that output the selected output values ​​as a candidate basis for prediction. In this case, a predetermined number of input data can be identified as candidates for the basis for prediction. Note that the information processing device 10 may also identify the input data associated with the output node that output a contribution degree (output value) equal to or greater than a predetermined value as a candidate basis for prediction. The learning model 12Ma of this embodiment may be configured to have a single output node that outputs information (e.g., an input node number) indicating the input data with the highest contribution degree, instead of having multiple output nodes that output the contribution degrees for each input data.

[0046] FIG. 8 is a flowchart showing an example of the recovery action presentation process procedure according to the second embodiment, and FIG. 9 is an explanatory diagram showing an example screen. The process shown in FIG. 8 is the process shown in FIG. 5 with step S31 added between steps S27 and S28. The same steps as those in FIG. 5 will not be described. The control unit 11 of the information processing device 10 according to this embodiment performs the same process as steps S21 to S27 in FIG. 5. As a result, in this embodiment, when a patient ID and medication information are input via the input screen, a recovery action to be performed on the patient is predicted based on the medication information and vital data, and advice information for notifying the predicted recovery action is generated. Note that in this embodiment, in step S26, the control unit 11 predicts a recovery action based on the output value from the learning model 12Ma and identifies input data that serve as the basis for the prediction result. For example, the control unit 11 selects a predetermined number (e.g., three) of contributions from the explanatory data (contributions of each input data) output from the learning model 12Ma in descending order, and identifies the input data associated with the output node that output the selected contributions as candidate basis for the prediction result.

[0047] After processing step S27, the control unit 11 generates prediction basis information for notifying the input data that forms the basis of the prediction result identified in step S26 (S31). For example, if a drug name has been identified as the input data that forms the basis of the prediction result, the control unit 11 generates a message (e.g., "Basis of recommended action: Drug A") notifying the drug name as the basis of the prediction result (recommended recovery action). Also, if blood pressure has been identified as the input data that forms the basis of the prediction result, the control unit 11 generates a message notifying the blood pressure as the basis of the prediction result (recommended recovery action). At this time, the control unit 11 may also determine whether the blood pressure has risen above the normal range or fallen below the normal range based on the vital data (blood pressure data) acquired in step S25, and notify the change in blood pressure as the basis of the prediction result.

[0048] Then, the control unit 11 displays an advice screen as shown in FIG. 9 on the display unit 15 based on the advice information generated in step S27 and the prediction basis information generated in step S31 (S28). The advice screen shown in FIG. 9 presents input data that contributed to the prediction of the recommended recovery action in addition to the contents of the advice screen shown in FIG. 6B. In the example shown in FIG. 9, three vital data items, namely, blood pressure, pulse rate, heart rate, and respiratory rate, are presented as the input data that contributed to the prediction result in descending order of contribution. Note that the contribution of each input data item may be displayed in association with each input data item. In this case, the contribution of each input data item to the prediction result can also be understood. The doctor selects an appropriate recovery action and implements the recovery action for the patient by checking the recommended recovery action on the advice screen and the input data that contributed to the prediction of the recovery action. In this embodiment, the doctor is presented with information on which of the medication information and vital data contributed to the prediction result (prediction of the recovery action), allowing the doctor to select a more appropriate recovery action.

[0049] In this embodiment, the learning model 12Ma is not limited to the configuration shown in FIG. 7 and can be configured to input various information contained in the electronic medical record data, such as height and weight, medical history, examination information, treatment information, surgery information, medication history, and comments entered by a doctor or the like. In this case, the learning model 12Ma can be configured to determine the recovery procedure to be performed based on not only medication information and vital data but also various information contained in the electronic medical record data, and to output the contribution of each input data to the determination result (prediction result). Note that the input electronic medical record data to the learning model 12Ma may also be time-series data. For example, the learning model 12Ma can be configured to input the patient's electronic medical record data for the most recent day or the most recent week.

[0050] In this embodiment, the same effects as those of the first embodiment described above can be obtained. Furthermore, in this embodiment, not only is the recovery procedure to be performed on the patient predicted, but the input data that serves as the basis for the prediction result can also be presented. Therefore, doctors, nurses, etc. can determine whether the presented input data and recovery procedure are appropriate, and if appropriate, adopt and perform the recovery procedure on the patient, thereby enabling early recovery of the patient's condition. In this embodiment, the modified examples described in the first embodiment described above can also be applied as appropriate.

[0051] (Embodiment 3) In the above-described first and second embodiments, if the patient's condition worsens during or after drug administration, the configuration is such that recovery measures to be taken by the patient are presented. In this embodiment, an information processing device is described that automatically determines whether or not the patient's vital data has become abnormal (whether or not the condition has worsened) based on the patient's vital data during or after drug administration, and if it is determined that the patient has become abnormal, it presents recovery measures to be taken by the patient. The information processing device of this embodiment has the same configuration as the information processing device 10 of embodiment 1, and therefore a detailed description of the configuration will be omitted.

[0052] FIG. 10 is a flowchart showing an example of a recovery action presentation process procedure according to the third embodiment, and FIG. 11 is an explanatory diagram showing an example screen. The process shown in FIG. 10 is the process shown in FIG. 5 with steps S41 to S42 added instead of steps S21 to S22. Explanation of the same steps as in FIG. 5 will be omitted. The information processing device 10 of this embodiment is connected to a monitor device (not shown) that measures, for example, a patient's vital data (blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, body temperature, etc.), determines whether the measured data is within a normal range, and outputs an alert if the measured data deviates from the normal range. In this case, the control unit 11 of the information processing device 10 can be configured to acquire an alert (vital data deviating from the normal range) detected by the monitor device from the monitor device. Note that the monitor device may be configured to output the measured vital data of the patient to the information processing device 10. In this case, the control unit 11 of the information processing device 10 determines whether the vital data acquired from the monitor device is within a normal range.

[0053] The control unit 11 of the information processing device 10 of this embodiment determines whether the patient's vital data has become abnormal depending on whether an alert has been received from a monitoring device monitoring the patient's vital data during or after medication (S41). If it is determined that the vital data is not abnormal (S41: NO), that is, if it is determined that the vital data is normal, the control unit 11 waits. If it is determined that the vital data has become abnormal (S41: YES), the control unit 11 generates an alert screen notifying the patient of the abnormal vital data and displays it on the display unit 15 (S42). FIG. 11A shows an example of the alert screen, which displays the patient ID of the patient being monitored and the vital data determined to be abnormal. In the example shown in FIG. 11A, the alert screen notifies the user that blood pressure has risen above the normal range. Similar to the input screen shown in FIG. 6A, the alert screen also has input fields for medication information (drug name, dosage, and administration rate) and an OK button.

[0054] After displaying the alert screen shown in FIG. 11A, the control unit 11 executes the processes from step S23 onward. That is, similar to the process shown in FIG. 5, the control unit 11 acquires medication information and displays it in each input field. When the OK button on the alert screen is operated, the control unit 11 acquires the patient's vital data from the electronic medical record DB 12a. Then, based on the acquired medication information and vital data, the control unit 11 identifies a recovery procedure using the learning model 12M and displays an advice screen on the display unit 15 that presents the identified recovery procedure. In this embodiment, an advice screen such as that shown in FIG. 11B is displayed. The advice screen shown in FIG. 11B notifies the patient of vital data that has become abnormal, in addition to the display contents of the advice screen shown in FIG. 6B. Such advice screen (display information) can present the patient's vital data that is not within the normal range and can also suggest recovery procedures to bring the vital data back into the normal range.

[0055] In the above-described process, the process of determining whether the patient's vital data is normal may be performed using a learning model trained by machine learning. For example, a learning model trained using an algorithm such as CNN, RNN, or LSTM can be used to output information indicating the patient's condition that changes due to medication when medication information regarding a drug being or has been administered to a patient and vital data from before, during, or after medication are input. In this case, the control unit 11 inputs the medication information regarding the drug being or has been administered and the vital data acquired from the monitoring device into the learning model, and based on the output information from the learning model, it becomes possible to determine the patient's current condition or predict their future condition.

[0056] FIG. 12 is an explanatory diagram showing an example of the configuration of a learning model used to determine a patient's condition. The learning model shown in FIG. 12 is trained to input medication information (drug name, dosage, and administration rate) of a drug being or has been administered to a patient and vital data of the patient during or after medication (blood pressure, pulse rate, heart rate, respiratory rate, electrocardiogram, body temperature, etc.), perform a calculation to determine the patient's condition based on the input data, and output the calculation result (discrimination probability for each condition). Note that the vital data input to the learning model may be time-series vital data from before, during, or after medication. The learning model shown in FIG. 12 may also be configured to input one or more pieces of information, such as test information, treatment information, surgery information, medication history, and comments entered by a doctor or the like, included in electronic medical record data, in addition to medication information and vital data. In addition, the electronic medical record data input into the learning model may be time-series data, such as time-series data (test information) of test results such as blood tests or urine tests performed regularly, or time-series data (treatment information) regarding various treatments performed regularly.

[0057] The learning model shown in FIG. 12 has multiple output nodes, and each output node outputs a discrimination probability for the state associated with it. For example, output node 0 outputs the probability that the patient's condition should be determined to be good, output node 1 outputs the probability that the blood pressure should be determined to be above the normal range, and output node 2 outputs the probability that the blood pressure should be determined to be below the normal range. The output value of each output node is, for example, a value between 0 and 1.0, and the sum of the discrimination probabilities output from each output node is 1.0. The state assigned to each output node can be a change in state that may occur due to medication.

[0058] The learning model shown in FIG. 12 is generated by machine learning an untrained learning model using training data including training medication information and vital data, and information indicating a patient's condition that may occur during or after medication (correct labels). When training medication information and vital data are input, the learning model learns so that the output value from the output node corresponding to the condition indicated by the correct label approaches 1.0 and the output values ​​from other output nodes approach 0.0. In the learning process, the learning model performs calculations based on the input medication information and vital data to calculate output values ​​from each output node. The learning model then compares the calculated output value of each output node with a value corresponding to the correct label (0 or 1) and optimizes parameters used in the calculation process so that each output value approximates the value corresponding to the correct label. Here, the optimized parameters are, for example, weights between neurons in the learning model, and the optimization method is not particularly limited. As a result, when a patient's medication information and vital data are input, a learning model trained to determine the patient's current condition or a condition that may occur in the future is obtained. The learning model may be trained by another learning device. The trained learning model generated by training on the other learning device is downloaded from the learning device to the information processing device 10 via a network or a portable storage medium 1a, for example, and stored in the storage unit 12.

[0059] When using a learning model such as that shown in FIG. 12, the control unit 11 of the information processing device 10 identifies the state associated with the output node that outputs the largest output value (discrimination probability) among the output values ​​from each output node as the patient's state. The control unit 11 may also select a predetermined number (e.g., three) of output values ​​from each output node in descending order of magnitude, and identify the state associated with the output node that output the selected output value as a candidate for the patient's state. When the patient's state is determined using the learning model shown in FIG. 12, the control unit 11 determines whether the patient's vital data is abnormal based on the predicted state in step S41 of FIG. 10. If the control unit 11 determines that the patient's state is abnormal, it displays an alert screen notifying the user of the abnormal vital data state (S42). Even with this configuration, the same processing as in the present embodiment described above is possible, and the same effects can be obtained.

[0060] This embodiment can achieve the same effects as the above-described embodiments. Furthermore, this embodiment can automatically determine whether a patient's condition is normal. Therefore, when an abnormal condition is automatically determined, the abnormal condition is notified and advice on recovery measures to recover from the abnormal condition can be provided. Therefore, this embodiment can detect a deterioration in a patient's condition early and provide advice on recovery measures to be implemented, thereby supporting doctors who need to make appropriate decisions in emergency situations or situations where support from a doctor with extensive experience and knowledge is not available.

[0061] The configuration of this embodiment is applicable to the information processing device 10 of embodiments 1 and 2, and similar effects can be obtained even when applied to the information processing device 10 of embodiments 1 and 2. Furthermore, the modified examples described in the above-mentioned embodiments can also be applied to this embodiment as appropriate.

[0062] (Embodiment 4) The following describes an information processing device that processes to receive instructions from a doctor to change the contents of the recovery measures presented when the patient's condition worsens during or after medication. The information processing device of this embodiment has the same configuration as the information processing device 10 of the first embodiment, so detailed description of the configuration will be omitted.

[0063] Fig. 13 is a flowchart showing an example of the recovery action presentation processing procedure of the fourth embodiment, and Fig. 14 is an explanatory diagram showing an example screen. The processing shown in Fig. 13 is the processing shown in Fig. 5 with steps S51 to S56 added after step S28. Explanation of the same steps as in Fig. 5 will be omitted. Note that steps S21 to S27 in Fig. 5 are not shown in Fig. 13.

[0064] The control unit 11 of the information processing device 10 of this embodiment performs the same processes as steps S21 to S28 in Fig. 5. As a result, also in this embodiment, when a patient ID and medication information are input via the input screen, a recovery measure to be taken on the patient is predicted based on the medication information and vital data, and an advice screen notifying the predicted recovery measure is displayed on the display unit 15. The control unit 11 of this embodiment displays an advice screen as shown in Fig. 14A. The screen shown in Fig. 14A has the same configuration as the advice screen shown in Fig. 6B, and the presented recovery measures are displayed so as to be selectable. Therefore, the control unit 11 accepts the selection of one of the recovery measures by operating the advice screen via the input unit 14. The example shown in Fig. 14A shows a state in which the recovery measure "Recommended 2" has been selected.

[0065] The control unit 11 determines whether a selection for any of the recovery measures displayed on the advice screen has been accepted (S51). If it is determined that a selection for a recovery measure has not been accepted (S51: NO), the control unit 11 waits until a selection is accepted. Note that if the close button on the advice screen is operated, the control unit 11 ends the series of processes. If the control unit 11 determines that a selection for any of the recovery measures has been accepted (S51: YES), the control unit 11 displays a change acceptance screen on the display unit 15 to accept an instruction to change the selected recovery measure (S52). FIG. 14B shows an example of the change acceptance screen. The screen shown in FIG. 14B displays the recovery measure selected on the screen shown in FIG. 14A, and has input fields where the content of the recovery measure can be changed. In the example shown in FIG. 14B, input fields are provided for "antagonist A1," "dose B1," and "administration rate C1," and the corresponding values ​​are displayed in each input field. The screen shown in Fig. 14B also has a change button for issuing an instruction to change the information entered in the input field, and a close button for issuing an instruction to end the display of the change acceptance screen. The doctor selects a recovery procedure to be performed on the patient by referring to the recovery procedure presented on the advice screen shown in Fig. 14A, and if the doctor wishes to change part of the content of the selected recovery procedure, he or she inputs the details of the correction to that part of the content of the recovery procedure via the change acceptance screen shown in Fig. 14B.

[0066] The control unit 11 accepts changes to the information displayed in the input fields on the change acceptance screen through operations via the input unit 14 (S53). When the control unit 11 accepts changes to any of the input fields, it displays the accepted changes in the input fields and updates them. The control unit 11 determines whether the change button on the change acceptance screen has been operated (S54), and if it determines that it has not been operated (S54: NO), it returns to the processing of step S53 and continues accepting changes to the input fields on the change acceptance screen.

[0067] When the control unit 11 determines that the change button has been operated (S54: YES), it acquires the medication information acquired in step S23, the vital data acquired in step S25, and the contents of the recovery procedure after the change including the changes accepted in step S53 (S55). Then, the control unit 11 generates training data by assigning a correct answer label indicating the recovery procedure after the change to the acquired medication information and vital data, and stores the generated training data in the storage unit 12 (S56). For example, the control unit 11 stores the generated training data in a DB for training data provided in the storage unit 12.

[0068] The above-described process can generate training data with the revised recovery action as the correct label when the doctor modifies the recovery action recommended when a patient's condition worsens due to medication. Therefore, by retraining the learning model 12M using such training data, the doctor's corrected recovery action can be fed back to the learning model 12M. This allows the generation of a learning model 12M that reflects the recovery action actually adopted by the doctor. In the above-described process, whether or not to use the training data using the corrected recovery action as training data can be determined based on whether the patient's condition has recovered to a normal state after the recovery action. The doctor determines whether the vital signs of the patient who underwent the recovery action have returned to a normal range and registers the determination result with the training data stored in the training data DB. In this case, when performing a learning process for the learning model 12M using the training data stored in the training data DB, the control unit 11 of the information processing device 10 can retrain using only the training data including the recovery action that resulted in the patient's recovery. In this case, the recovery action that was effective for the patient can be fed back to the learning model 12M.

[0069] This embodiment can achieve the same effects as the above-described embodiments. Furthermore, this embodiment can generate a learning model 12M that reflects the contents of recovery measures actually adopted by a doctor, and by using this learning model 12M, it is possible to provide advice on more appropriate recovery measures.

[0070] The configuration of this embodiment is applicable to the information processing device 10 of embodiments 1 to 3, and similar effects can be obtained even when applied to the information processing device 10 of embodiments 1 to 3. Furthermore, the modified examples described in each of the above-mentioned embodiments can also be applied to this embodiment as appropriate.

[0071] The embodiments disclosed herein are to be considered as illustrative in all respects and not restrictive. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the meaning and scope of the claims. [Explanation of symbols]

[0072] 10. Information processing equipment 11 Control section 12 Storage section 13 Communications Department 14 Input section 15 Display 12a Electronic medical record database 12M learning model 12Ma learning model

Claims

1. Acquire vital data of a patient who has become abnormal during or after administration of an injection, as well as the type, dosage, and administration rate of the injection; a learning model that has been trained to output, when the type, dosage, and administration rate of an injection and vital data are input, the probability of performing each of a plurality of recovery measures, including discontinuing administration of the injection, changing the dosage or administration rate of the injection, the type, dosage, and administration rate of an antagonist for the injection, and changing the injection, by inputting the acquired type, dosage, and administration rate of the injection and vital data, to obtain the probability for each of the plurality of recovery measures; and outputting advice information regarding a plurality of recovery measures to be taken on the patient who has become in the abnormal state based on the probability of each of the plurality of recovery measures. A program that causes a computer to perform a process.

2. obtaining patient information regarding the patient; the learning model is trained to output a probability that each of the plurality of recovery procedures should be performed when the type, dosage, and administration rate of the injection, vital data, and patient information are input; The acquired type, dosage, and administration rate of the injection, vital data, and patient information are input into the learning model to acquire the probability of performing each of the plurality of recovery procedures. The program according to claim 1 , which causes the computer to execute a process.

3. and outputting display information that displays vital data indicating an abnormal state and advice information regarding the plurality of recovery measures.

3. The program according to claim 1, which causes the computer to execute processing.

4. acquiring, from the learning model, information corresponding to the contribution of each piece of data input to the learning model to the output of the probability for the plurality of recovery actions; Based on the acquired information, among the data input to the learning model, data that has a high contribution to the output of the probability is output.

4. The program according to claim 1, which causes the computer to execute a process.

5. The advice information includes information about the type of injection, the dosage, and the administration rate.

5. The program according to claim 1.

6. Accepts instructions to change the output advice information regarding the recovery measures.

6. The program according to claim 1, which causes the computer to execute a process.

7. generating training data including the type, dosage, and administration rate of the injection drug input into the learning model, as well as vital data and information on recovery measures taken when the patient recovered; The learning model is retrained using the generated training data.

7. The program according to claim 1, which causes the computer to execute a process.

8. Acquire vital data of a patient who has become abnormal during or after administration of an injection, as well as the type, dosage, and administration rate of the injection; a learning model that has been trained to output, when the type, dosage, and administration rate of an injection and vital data are input, the probability of performing each of a plurality of recovery measures, including discontinuing administration of the injection, changing the dosage or administration rate of the injection, the type, dosage, and administration rate of an antagonist for the injection, and changing the injection, by inputting the acquired type, dosage, and administration rate of the injection and vital data, to obtain the probability for each of the plurality of recovery measures; and outputting advice information regarding a plurality of recovery measures to be taken on the patient who has become in the abnormal state based on the probability of each of the plurality of recovery measures. An information processing method in which processing is performed by a computer.

9. an acquisition unit that acquires vital data of a patient who has become abnormal during or after administration of an injection, as well as the type, dosage, and administration rate of the injection; a learning model that has been trained to output, when the type, dosage, and administration rate of an injection and vital data are input, the probability of performing each of a plurality of recovery measures, including discontinuing administration of the injection, changing the dosage or administration rate of the injection, the type, dosage, and administration rate of an antagonist for the injection, and changing the injection, by inputting the acquired type, dosage, and administration rate of the injection and vital data, to obtain the probability for each of the plurality of recovery measures; an output unit that outputs advice information regarding a plurality of recovery measures to be taken on the patient who has become in the abnormal state, based on the probability of each of the plurality of recovery measures; An information processing device comprising:

10. acquiring training data including the type, dosage, and administration rate of an injectable agent administered to a patient, vital data of the patient after administration of the injectable agent, and information indicating any of a plurality of recovery measures taken by the patient, including discontinuing administration of the injectable agent, changing the dosage or administration rate of the injectable agent, the type, dosage, and administration rate of an antagonist for the injectable agent, and changing the injectable agent; Using the acquired training data, a learning model is generated that outputs the probability of performing each of a plurality of recovery measures, including stopping administration of the injection, changing the injection dose or injection rate, the type, dosage and injection rate of an antagonist for the injection, and changing the injection, when the type, dosage, and injection rate of the injection and vital data are input. A model generation method in which processing is performed by a computer.

Images