Program, information processing method, and information processing device

Abstract

The present invention provides a program and the like capable of supporting processing for determining prescription details of a drug to be administered to a subject on the basis of the edema condition of the subject. This computer stores, in a storage unit, prescription information of a drug prescribed to a subject and time-series data of an index indicating the degree of edema of the subject. Then, the computer outputs a first screen for displaying, side by side in association with each date and time, information on the drug administered at the date and time and an index indicating the degree of edema measured at the date and time.

Description

Program, information processing method, and information processing device

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

[0002] Body composition analyzers are used that measure the resistance value (bioelectric impedance) of a subject's living body by passing a weak current through the living body, and estimate values ​​related to the subject's body composition based on the measured bioelectric impedance and physical information such as the subject's age, sex, height, weight, etc. Patent Document 1 discloses a device that calculates the degree of edema in a target part of the subject's body according to the bioimpedance of the target part, and displays an edema stage that indicates the progression of edema symptoms.

[0003] JP 2013-233357 A

[0004] When treating edema symptoms, the type and dosage of medication must be determined taking into account the progression and degree of improvement of the edema symptoms, as well as the progression of the edema. In particular, edema symptoms caused by congestive heart failure require early treatment (water removal), and the type, dosage, and route of administration (intravenous or oral) of the medication to be administered must be determined at the appropriate time. Making an incorrect decision at this stage can have a negative impact on the treatment course and prognosis. Insufficient water removal can result in a prolonged hospital stay due to poor improvement in edema symptoms, or residual edema or congestion at the time of discharge, which can lead to short-term readmission. On the other hand, excessive water removal in a short period of time can place strain on organs such as the kidneys, adversely affecting long-term prognosis. The device disclosed in Patent Document 1 can be used to assess the progression of edema symptoms. However, prescriptions based on the progression and improvement of edema symptoms are determined based on the knowledge and experience of medical professionals, such as physicians, making it difficult to determine the appropriate prescription.

[0005] In one aspect, an object of the present invention is to provide a program or the like capable of assisting in the process of determining the prescription contents of a drug to be administered to a subject based on the subject's edema state.

[0006] (1) The present disclosure is a program that causes a computer to execute a process of storing prescription information of a drug prescribed to a subject in a memory unit, storing time series data of an index indicating the degree of edema of the subject in the memory unit, and outputting a first screen that displays information on the drug administered at each date and time and an index indicating the degree of edema measured at each date and time side by side, in association with the date and time.

[0007] (2) It is preferable that the program of (1) above further causes the computer to execute a process of generating the first screen that creates a graph showing the time series change of the index indicating the degree of edema, with the date and time shown on the first axis and the value of the index on the second axis, based on the time series data of the index indicating the degree of edema, and displays the graph alongside information on the drugs administered at each date and time, corresponding to each date and time shown on the first axis of the graph.

[0008] (3) It is preferable that the program of (2) above further causes the computer to execute a process of adding a mark indicating a normal range set for the index indicating the degree of edema onto the graph in the first screen.

[0009] (4) It is preferable that the program of (2) or (3) above further causes the computer to execute the following process: identify the date and time when the administration method of the drug was changed from intravenous injection to oral administration based on the prescription information or based on date and time identification information received from the user; add a line indicating the identified date and time to the drug information and the graph in the first screen; and add different backgrounds to the area before the date and time indicated by the line and the area after the date and time indicated by the line.

[0010] (5) It is preferable that the program described in any of (2) to (4) above further causes the computer to execute a process of identifying the date and time when the administration method of the drug was changed from intravenous injection to oral administration based on the prescription information or based on date and time identification information received from the user, and adding a line or mark indicating the identified date and time on the graph in the first screen.

[0011] (6) It is preferable that the program described in any of (2) to (5) above further causes the computer to execute a process of identifying the date and time when the administration method of the drug was changed from intravenous injection to oral administration based on the prescription information or based on date and time identification information received from the user, and adding different backgrounds to the area before the identified date and time and the area after the identified date and time on the graph in the first screen.

[0012] (7) It is preferable that the program described in any of (2) to (6) above further causes the computer to execute a process of calculating the dosage of the drug administered at each time interval based on the prescription information and the time interval indicating the measurement interval of the index indicating the degree of edema, and generating the first screen that corresponds to the first axis of the graph and displays information on the drug administered at each time interval.

[0013] (8) It is preferable that the program described in any of (2) to (7) above further causes the computer to execute a process of inputting the prescription information of the subject and the time series data of the index indicating the degree of edema stored in the memory unit, and the subject's physical data into a learning model that has been trained to output an index indicating the subject's future degree of edema when the prescription information of the subject, time series data of the index indicating the degree of edema, and the subject's physical data are input, to obtain an index indicating the subject's future degree of edema, and to add the obtained index indicating the future degree of edema to the graph in the first screen in association with a future date and time.

[0014] (9) It is preferable that the program of (8) above further causes the computer to execute a process of adding an index indicating the degree of future edema in a manner different from other values ​​of the index in the graph.

[0015] (10) It is preferable that the program described in any of (1) to (9) above outputs a second screen that displays prescription information for multiple subjects, including the type of prescribed drug, dosage, and administration method, as well as the number of days since the administration method was changed from intravenous injection to oral administration, in association with an index indicating the degree of edema and / or the trend of change in said index, and further causes the computer to execute a process of accepting the selection of one of the multiple subjects via the second screen and outputting the first screen regarding the selected subject.

[0016] (11) In the program described in any one of (1) to (10) above, the index indicating the degree of edema is preferably an edema index calculated as the ratio of extracellular fluid volume (ECW) to total body water (TBW).

[0017] (12) The present disclosure also provides an information processing method in which a computer executes a process of storing prescription information of a drug prescribed to a subject in a storage unit, storing time series data of an index indicating the degree of edema of the subject in the storage unit, and outputting a first screen that displays information of the drug administered at each date and time and an index indicating the degree of edema measured at each date and time side by side, in association with the date and time.

[0018] (13) The present disclosure also provides an information processing device having a control unit, wherein the control unit stores prescription information of a drug prescribed to a subject in a memory unit, stores time series data of an index indicating the degree of edema of the subject in the memory unit, and outputs a first screen that displays information on the drug administered at each date and time and an index indicating the degree of edema measured at each date and time side by side, in association with the date and time.

[0019] In one aspect, the method can assist in determining a medication regimen to administer to a subject based on the subject's edema status.

[0020] FIG. 1 is an explanatory diagram showing an example of the configuration of an information processing system. FIG. 2 is a block diagram showing an example of the configuration of an information processing device and an electronic medical record server. FIG. 3 is an explanatory diagram showing an example of the record layout of an electronic medical record DB. FIG. 4 is a flowchart showing an example of a display processing procedure for prescription contents and edema rate. FIG. 5 is an explanatory diagram showing an example of a screen. FIG. 6 is an explanatory diagram showing an example of a screen. FIG. 7 is an explanatory diagram showing an example of a screen. FIG. 8 is an explanatory diagram showing a modified example of a patient details screen. FIG. 9 is an explanatory diagram showing an example of the configuration of a learning model. FIG. 10 is a flowchart showing an example of a processing procedure for estimating a future edema rate. FIG. 11 is an explanatory diagram showing an example of a screen. FIG. 12 is an explanatory diagram showing an example of a screen. FIG. 13 is a flowchart showing an example of a processing procedure for estimating a future edema rate according to embodiment 3.

[0021] Hereinafter, a program, an information processing method, and an information processing device according to the present disclosure will be described in detail with reference to the drawings illustrating embodiments thereof.

[0022] (Embodiment 1) In this embodiment, an information processing system is described that efficiently presents information to be viewed by a medical professional such as a doctor when determining the prescription contents of a drug to be administered to a subject (hereinafter referred to as a patient) based on the degree of the patient's edema symptoms. Note that in this embodiment, the drug to be prescribed includes a diuretic, since the purpose is to treat the patient's edema symptoms.

[0023] FIG. 1 is an explanatory diagram showing an example configuration of an information processing system. The information processing system of this embodiment includes an information processing device 10, a body composition analyzer 20, and an electronic medical record server 30. The body composition analyzer 20 and the electronic medical record server 30 are configured to directly transmit and receive information via wired or wireless communication via a cable. The information processing device 10 and the electronic medical record server 30 are connected via a network N. The network N may be the Internet or a public communication line, or may be a local area network (LAN) established within a facility such as a medical institution or testing institution where the information processing system is installed. The body composition analyzer 20 may be configured to be connectable to the network N. In this case, the body composition analyzer 20 and the electronic medical record server 30 transmit and receive information via the network N. In this embodiment, the patient's body composition values ​​measured (measured) by the body composition analyzer 20 (hereinafter referred to as body composition information) are directly transmitted to the electronic medical record server 30. However, the body composition information measured by the body composition analyzer 20 may also be transmitted to the information processing device 10 and then transmitted from the information processing device 10 to the electronic medical record server 30.

[0024] The body composition analyzer 20 is a measuring device that measures bioelectrical impedance, which is the resistance value of a patient's living body, using bioelectrical impedance analysis (BIA) and measures (estimates) the patient's body composition based on the measured bioelectrical impedance. The body composition analyzer 20 has electrodes 20a and 20b attached to the patient's wrists and ankles. The body composition analyzer 20 measures the patient's bioelectrical impedance based on the current flowing between the electrodes 20a and 20b and the voltage between the electrodes 20a and 20b, as indicated by the dashed lines in FIG. 1 . Each of the electrodes 20a and 20b on the body composition analyzer 20 may be provided with one electrode or two electrodes. Specifically, the body composition analyzer 20 applies a current to one of the electrodes 20a and 20b attached to the patient and measures the patient's bioelectrical impedance based on the resulting voltage between the electrodes 20a and 20b. The method for measuring bioelectrical impedance is based on a known BIA, and therefore a detailed description thereof will be omitted.

[0025] The body composition monitor 20 has a calculation unit (not shown) that calculates the patient's body composition information from the bioelectrical impedance measured using the electrodes 20a and 20b. The body composition monitor 20 also has a memory unit (not shown) that stores each patient's physical data, including their age, gender, height, and weight. The calculation unit of the body composition monitor 20 calculates each patient's body composition information based on the measured bioelectrical impedance and each patient's physical data. The body composition monitor 20 also stores the measured bioelectrical impedance and calculated body composition information in the memory unit, along with a timestamp representing the current date and time (measurement date and time). The body composition monitor 20 also has a communication unit (not shown) that transmits and receives information to the electronic medical record server 30. The body composition monitor 20 transmits the patient's body composition information calculated by the calculation unit, along with the timestamp, to the electronic medical record server 30, where it is stored in the electronic medical record database 32a (see FIG. 2 ). The body composition monitor 20 transmits the patient's body composition information to the electronic medical record server 30 in accordance with instructions from the user received via an input unit (not shown) or instructions from the electronic medical record server 30 to which it is connected in communication.

[0026] The information processing device 10 is an information terminal used by staff (medical professionals) at a medical institution, etc. The information processing device 10 is a device capable of various information processing and sending and receiving information, and is composed of a personal computer, tablet terminal, smartphone, etc. The electronic medical record server 30 is an information processing device capable of various information processing and sending and receiving information, and is composed of a server computer, personal computer, etc. The electronic medical record server 30 may be a multi-computer consisting of multiple computers, or may be a virtual machine virtually constructed within a single device using software. The electronic medical record server 30 may also be a local server installed within a medical institution, or a cloud server connected to the medical institution via the Internet.

[0027] FIG. 2 is a block diagram showing an example configuration of the information processing device 10 and the electronic medical record server 30. 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, and the like, which are interconnected via a bus. The control unit 11 has one or more processors, such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), or a GPU (Graphics Processing Unit). The control unit 11 appropriately executes a program 12P stored in the storage unit 12 to perform various information processing, control processing, and the like that the information processing device 10 should perform. Note that if the control unit 11 has multiple processors, each process may be performed by a different processor.

[0028] 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 program 12P (a program product, a computer program) executed by the control unit 11 and various data required for executing the program 12P. The storage unit 12 also temporarily stores data generated when the control unit 11 executes the program 12P.

[0029] The communication unit 13 has a communication module for connecting to the network N via wired or wireless communication, and transmits and receives information to other devices via the network N. The input unit 14 accepts operation input by a user (e.g., a medical professional) 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 in accordance with instructions from the control unit 11. A part of the input unit 14 and the display unit 15 may be an integrated touch panel. Note that the input unit 14 and the display unit 15 are not essential, and the information processing device 10 may be configured to accept operations via a connected terminal device and output information to be displayed to an external display device.

[0030] The reading unit 16 reads information stored in a portable storage medium 10a, such as a CD (Compact Disc), a DVD (Digital Versatile Disc), a USB memory, or an SD (Secure Digital) card. The program 12P and various data stored in the storage unit 12 may be read by the control unit 11 from the portable storage medium 10a via the reading unit 16 and stored in the storage unit 12. The program 12P and various data may be written to the storage unit 12 during the manufacturing stage of the information processing device 10, or may be downloaded by the control unit 11 from another device via the communication unit 13 and stored in the storage unit 12.

[0031] The information processing device 10 is not limited to a single computer, but may be a multi-computer including multiple computers. Furthermore, the information processing device 10 may be a virtual machine virtually constructed within a single device by software, or may be a cloud server. In the following description, the information processing device 10 is described as a single computer. Furthermore, the program 12P may be deployed and executed on a single computer or at a single site, or may be distributed across multiple sites and deployed to be executed on multiple computers interconnected by a network N.

[0032] The electronic medical record server 30 includes a control unit 31, a memory unit 32, a communication unit 33, an input unit 34, a display unit 35, a reading unit 36, etc., and these units are interconnected via a bus. The control unit 31, memory unit 32, communication unit 33, input unit 34, display unit 35, and reading unit 36 ​​of the electronic medical record server 30 have the same configurations as the units 11 to 16 of the information processing device 10, and therefore their description will be omitted. The memory unit 32 of the electronic medical record server 30 stores an electronic medical record DB 32a in addition to the program 32P executed by the control unit 31. The input unit 34 and display unit 35 are not essential to the electronic medical record server 30; the server may be configured to accept operations via a connected computer or to output information to be displayed to an external display device.

[0033] FIG. 3 is an explanatory diagram showing an example of the record layout of the electronic medical record DB 32a. The electronic medical record DB 32a is a database that stores electronic medical record data (medical records) of patients who use medical institutions, etc. The electronic medical record DB 32a shown in FIG. 3 includes a patient ID column, a name column, a patient information column, a prescription information column, a body composition information column, etc. 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 name column stores the patient's name in association with the patient ID. The patient information column stores patient information related to the patient in association with the patient ID. The patient information includes, for example, attribute information including the patient's age and gender, personal information including the date of birth and contact information, physical data including the patient's height and weight, diagnosis information including the diagnosis, start date of hospitalization, and hospital room. The prescription information column stores prescription information of drugs prescribed for the patient's symptoms being treated in association with the patient ID. The prescription information includes the prescription date and time and prescription content. The prescription contents include information on the drug administration period, drug type, dosage, and administration method and route (e.g., intravenous injection or oral administration). The body composition information column stores body composition information (body composition measurement data) measured using the body composition analyzer 20 in association with a patient ID. The body composition information is time-series data including the measurement date and time and various body composition values ​​measurable by the body composition analyzer 20. In this embodiment, the body composition information includes at least an edema index (extracellular water ratio) as an index indicating the degree of edema, and may further include fat-free mass (FFM), fat mass, body fat percentage, Fat Free Mass Index (FFMI), phase angle, total body water (TBW), extracellular water (ECW), intracellular water (ICW), hydration rate, muscle mass, skeletal muscle mass, skeletal muscle mass index (SMI), internal fluid ratio, external fluid ratio, internal fluid / external fluid ratio, etc. The edema rate is a value calculated as the ratio of extracellular fluid volume (ECW) to total body water (TBW).

[0034] In addition to the information shown in FIG. 3 , the electronic medical record DB 32 a may store various types of patient information, such as vital data (e.g., body temperature, blood pressure, heart rate, respiratory rate, oxygen saturation), medical history (medical history), examination information (e.g., blood tests, urine tests), treatment information (e.g., treatments), surgical information (e.g., surgery), and medication history. The examination information may also include the maximum and minimum diameters of the inferior vena cava (IVC) and the respiratory variation rate of the IVC obtained by echocardiography, as well as medical images such as X-ray images, ultrasound images, computed tomography (CT) images, magnetic resonance imaging (MRI) images, and electrocardiograms. Furthermore, the examination information may also include test values ​​(e.g., pulmonary artery wedge pressure, CVP, cardiac output, cardiac index) obtained by a thermodilution catheter, and test values ​​(e.g., LVEF, E / e′, valvular disease test values) obtained by echocardiography.

[0035] The following describes a process for presenting a medical professional with the prescription details and edema rate for each patient. FIG. 4 is a flowchart illustrating an example of the procedure for displaying the prescription details and edema rate, and FIGS. 5 to 6B are explanatory diagrams showing example screens. In the information processing system of this embodiment, a medical professional measures a patient's body composition information using a body composition analyzer 20, then connects the body composition analyzer 20 to the electronic medical record server 30 via communication. The medical professional transmits the patient ID, measurement date and time, and measurement data (body composition information) stored in the body composition analyzer 20 to the electronic medical record server 30. The electronic medical record server 30 stores the body composition information (measurement date and measurement data) received from the body composition analyzer 20 in the electronic medical record DB 32a, correlating it with the received patient ID. An instruction to transmit data from the body composition analyzer 20 to the electronic medical record server 30 may be received via an input unit of the body composition analyzer 20 or via an input unit 34 of the electronic medical record server 30. The medical staff may also specify a patient and issue an instruction to send the data of the specified patient from the body composition analyzer 20 to the electronic medical record server 30, and this operation may be performed from a viewing page for each patient on the electronic medical record server 30. Furthermore, the medical staff may issue an instruction to send the data of all patients stored in the body composition analyzer 20 from the body composition analyzer 20 to the electronic medical record server 30.

[0036] When the control unit 11 of the information processing device 10 receives an instruction to view the contents of the electronic medical record DB 32a, specifically, an instruction to view the edema rate of each patient, the control unit 11 performs the following process. The control unit 11 acquires each piece of patient information stored in the electronic medical record DB 32a (S11) and generates a patient list screen (second screen) based on the acquired information and displays it on the display unit 15 (S12). For example, the control unit 11 reads the patient ID, name, gender, age, current condition, prescription information, and edema rate for each patient cared for by the medical professional who is the viewer from the electronic medical record DB 32a. Note that the control unit 11 may read the most recent information as the prescription information, or the edema rate for a predetermined period up to the most recent period as the edema rate. Then, as shown in FIG. 5, the control unit 11 generates a patient list screen that displays each patient's information as a single record data. In the example of FIG. 5, the control unit 11 displays information indicating whether the patient is currently hospitalized in an ICU (Intensive Care Unit) or a general ward as the current condition. The current status may include not only the type of room the patient is in (ICU or general ward) but also symptoms of the patient that require attention (e.g., shortness of breath). The control unit 11 also displays the most recent prescription details (type of drug, dosage, administration method, number of days since the administration method was changed from intravenous injection to oral administration, etc.) as prescription information. Furthermore, the control unit 11 displays the most recent edema rate and a mark indicating the trend in the edema rate (maintaining the current state, increasing, decreasing, etc.) over a specified period of time (e.g., several days, several hours) up to the most recent period.

[0037] The patient list screen is configured to accept a selection of a patient for a record data item when a predetermined operation (e.g., double-clicking a mouse) is performed on the record data item. The control unit 11 determines whether a patient selection has been accepted through the predetermined operation on the patient list screen (S13). If it determines that a patient selection has not been accepted (S13: NO), the control unit 11 waits. If it determines that a patient selection has been accepted (S13: YES), the control unit 11 generates a detailed screen (first screen) displaying the prescription information and edema rate of the selected patient and displays it on the display unit 15, as shown in FIGS. 6A and 6B. Specifically, the control unit 11 obtains the prescription information and edema rate of the selected patient from the electronic medical record DB 32a of the electronic medical record server 30 (S14). Here, the control unit 11 obtains prescription information for the symptoms being treated and the edema rate during the treatment period.

[0038] The control unit 11 generates a graph (line graph) showing the time-series change in the edema rate based on the edema rate acquired from the electronic medical record DB 32a (S15). Region R2 in the screens shown in FIGS. 6A and 6B represents the graph showing the time-series change in the edema rate. The edema rate graph shows time on the horizontal axis (first axis) and the edema rate (Edema Index) on the vertical axis (second axis). For example, if the edema rate is measured once a day, the control unit 11 generates a graph showing the time-series change in the edema rate for each day, as shown in FIG. 6A. Note that the horizontal axis of the edema rate graph in FIG. 6A indicates the number of days of hospitalization, with the start date of hospitalization as Day 1. However, the measurement date or measurement date and time of each edema rate may also be used. Furthermore, if the edema rate is measured at predetermined time intervals (e.g., every few hours), the control unit 11 may generate a graph showing the time-series change in the edema rate for each predetermined time, with the horizontal axis showing the measurement date and time for each predetermined time.

[0039] The control unit 11 adds marks (markers) indicating the normal range set for the edema rate on the graph of the edema rate (S16). In the example of FIG. 6A, the normal range area on the graph of the edema rate is hatched. The normal range of the edema rate is, for example, 0.36 to 0.40, and is stored in the memory unit 12. Therefore, the control unit 11 can add marks indicating the normal range on the graph of the edema rate by reading the normal range of the edema rate from the memory unit 12. The normal range of the edema rate is not limited to the above range, and a normal range may be set according to, for example, the condition of the disease, or the configuration may be such that input is accepted.

[0040] Next, the control unit 11 calculates the prescription content corresponding to each measurement date and time of the edema rate based on the prescription information acquired from the electronic medical record DB 32a (S17). For example, if the edema rate is measured once a day, the control unit 11 calculates the prescription content for each day from the prescription information. For example, if the same drug is prescribed multiple times in a day, the control unit 11 calculates the total drug dosage and sets this as the drug dosage for that day. Furthermore, if the edema rate is measured at predetermined time intervals (e.g., every few hours), the control unit 11 calculates the prescription content for each predetermined time from the prescription information. For example, if the daily dosage of a certain drug is registered as prescription information, the control unit 11 calculates a single dosage by dividing the daily dosage by the number of times a day (24 hours) is divided into predetermined periods, and sets this as the drug dosage for that predetermined time. The dosage may also be displayed in terms of drug potency.

[0041] In this embodiment, the prescription content corresponding to each measurement timing is calculated based on the edema rate measurement timing as described above. However, the prescription content corresponding to each time at any time interval may also be calculated. For example, as shown in FIG. 6B , the prescription content corresponding to each time every 0.5 days (12 hours) may be calculated. In this case, the dosage of each drug in the prescription content on the screen of FIG. 6B is half the prescribed amount in the prescription content on the screen of FIG. 6A. Note that when the prescription content chart displays the prescription content at each time at any time interval, the control unit 11 identifies the edema rate at each time interval and creates a graph showing the change in the edema rate over time at each time interval in step S15. For example, if the edema rate is measured once a day and a graph is created showing the change over time in the edema rate every 0.5 days (12 hours), the same numerical value is used for the edema rate on days 0.5 and 1, the edema rate on days 1.5 and 2, the edema rate on days 2.5 and 3, the edema rate on days 3.5 and 4, and the edema rate on days 4.5 and 5, as shown in Figure 6B. As shown in Figures 6A and 6B, the time intervals indicated by the time axis can be changed arbitrarily, and the graph can be displayed at time intervals that make it easy for medical professionals to understand the patient's condition depending on the patient's symptoms, etc.

[0042] The control unit 11 then creates a chart showing the change in prescription content over time based on the prescription content corresponding to each measurement time (S18). Region R1 in the screen shown in FIGS. 6A and 6B represents the chart showing the change in prescription content over time. The prescription content chart has the same horizontal axis as the edema rate graph, and displays the prescription content corresponding to each time indicated by the horizontal axis. That is, the prescription content chart uses the time interval indicated by the horizontal axis of the edema rate graph (the interval between measurement dates and times) as the unit time, and displays the prescription content including the type of drug (first drug, second drug in FIGS. 6A and 6B), the dosage of each drug, and the administration method for each unit time. Note that, to simplify the prescription content chart, identical prescription content is displayed together. In the example of FIGS. 6A and 6B, the prescription content of the first drug on days 1 and 2 and the prescription content of the first drug on days 4 to 10 are displayed together. In the examples of Figures 6A and 6B, when the dosage and administration method are the same, the start time of the prescription is indicated by a circle, and a line is drawn for the duration of the prescription. Through the processing up to this point, a detailed screen is generated, as shown in Figures 6A and 6B, with region R1 displaying a chart of the prescription contents and region R2 displaying a graph of the edema rate. It is preferable that region R1 display the efficacy of each drug according to its estimated efficacy. Specifically, as shown in Figures 6A and 6B, the stronger the estimated efficacy, the higher the line indicating the duration of each prescription content may be positioned, or the color of the circle indicating the start time of each prescription content and the line indicating its duration may be changed depending on the strength of the estimated efficacy.

[0043] Next, the control unit 11 determines whether there is a transition timing for switching the administration method of any drug from intravenous injection to oral administration based on the acquired prescription information (S19). Specifically, the control unit 11 compares the administration method in the currently acquired prescription information with the administration method in the previously acquired prescription information, and if they differ, determines that there is a transition timing. Note that information indicating the timing of the administration method transition (date and time identification information) may be received by input from the user (healthcare professional) via the input unit 14. In this case, the control unit 11 determines whether there is a transition timing based on whether the user inputs the date and time identification information. If there is a transition timing (S19: YES), the control unit 11 identifies the date and time of the transition timing and adds a line L indicating the transition timing to the edema rate graph and the prescription content chart (S20). Note that if there is a transition timing based on the prescription information, the control unit 11 identifies the date and time of the transition timing from the prescription information. If there is a transition timing information input by the user, the control unit 11 identifies the date and time of the transition timing from the input information. In addition, the control unit 11 adds different backgrounds to the area before the transition timing (the area before the date and time indicated by line L) and the area after the transition timing (the area after the date and time indicated by line L) in the edema rate graph and the prescription content chart (S21). This makes it possible to display the edema rate graph and the prescription content chart with different backgrounds for each drug administration method. In the edema rate graphs in Figures 6A and 6B, the area during which the first drug was administered by intravenous injection is shown in dark gray, and the area during which the first drug was administered orally is shown in light gray.

[0044] If it is determined that there is no timing for transition (S19: NO), i.e., if the administration method of each drug is intravenous injection only or oral administration only, the control unit 11 adds a background according to the administration method (intravenous injection or oral administration) to the edema rate graph and the prescription content chart (S22). For example, if the administration method is intravenous injection only, the control unit 11 adds dark gray to the background of all the edema rate graphs and prescription content charts, and if the administration method is oral administration only, the control unit 11 adds light gray to the background of all the edema rate graphs and prescription content charts.

[0045] The control unit 11 generates a detailed screen, such as that shown in FIGS. 6A and 6B, for the patient selected via the screen of FIG. 5 through the above-described process, and displays the generated detailed screen (first screen) on the display unit 15 (S23). Through the above-described process, the screens shown in FIGS. 6A and 6B can display the prescription content and edema rate for each date and time shown on the horizontal axis side by side. For example, for a patient receiving intravenous administration of a drug (diuretic), if the edema rate is 0.40 or higher and is on the rise, the physician should consider changing the type of drug or increasing the dosage. Furthermore, if the edema rate is less than 0.40, the physician should consider changing the type of drug or reducing the dosage. If the edema rate is less than 0.40 but is on the rise, the physician should consider changing the type of drug or increasing the dosage. Furthermore, if the edema rate is less than 0.36 and is on the decline, the physician should consider switching to oral administration in addition to changing the type of drug or reducing the dosage. Furthermore, for patients receiving oral medication, if the edema rate reaches 0.40 or above and is on the rise, physicians should consider changing the medication type, increasing the dosage, or even returning to intravenous administration. Furthermore, if a patient receiving oral medication maintains an edema rate below 0.40, physicians should consider changing the medication type or reducing the dosage. If the edema rate remains within the normal range (e.g., 0.36 or higher and 0.40 or lower), physicians should consider whether discharge is possible and what prescriptions to prescribe after discharge. Thus, physicians need to appropriately modify prescriptions based on the patient's medication status, the severity of edema symptoms (edema rate), and trends in edema. By displaying the prescription and edema rate at each time point side-by-side on screens such as those shown in Figures 6A and 6B, physicians can efficiently present information that should be viewed when determining medication prescriptions. This, in turn, assists physicians in determining medication prescriptions.

[0046] The detailed screen is not limited to the configuration shown in Figures 6A and 6B. For example, the line L indicating the date and time of the transition timing may not be added, and different backgrounds may be added to the area before the transition timing and the area after the transition timing, or only the line L indicating the date and time of the transition timing may be added. Furthermore, instead of the line L indicating the date and time of the transition timing, a mark (marker) may be added to the location of the date and time of the transition timing on the horizontal axis, for example. Furthermore, the line L indicating the date and time of the transition timing may be added only to the graph of the edema rate, or different backgrounds may be added to the area before the transition timing and the area after the transition timing only to the graph of the edema rate.

[0047] FIG. 7 is an explanatory diagram showing a modified example of the patient details screen. Similar to the screens in FIGS. 6A and 6B , the screen in FIG. 7 includes an area R1 displaying a chart of prescription details and an area R2 displaying a graph of edema rates. It also includes an area R3 displaying graphs showing changes in body weight and urine volume over time, and an area R4 displaying graphs showing changes in the maximum and minimum IVC diameters and respiratory variation rate over time. The weight graph (line graph) shows time on the horizontal axis and weight on the vertical axis, while the urine volume graph (bar graph) shows urine volume per hour over time on the vertical axis. The IVC maximum and minimum diameter graph (line graph) shows time on the horizontal axis and IVC diameter on the vertical axis, while the IVC respiratory variation rate graph (bar graph) shows time on the horizontal axis and IVC respiratory variation rate over time on the vertical axis. The IVC diameter graph shows the change in maximum diameter over time with a white square, and the change in minimum diameter over time with a white triangle. The horizontal axes of the graphs of body weight and urine volume, and the horizontal axes of the graphs of IVC diameter and respiratory variation rate indicate the same unit of time as the horizontal axes of the prescription content chart and edema rate graph, and each graph shows body weight, urine volume, IVC diameter, and respiratory variation rate corresponding to the timing of edema rate measurement or each time at any time interval.

[0048] When creating the screen shown in FIG. 7 , the control unit 11 acquires the weight and urine volume of the selected patient from the electronic medical record DB 32a after processing step S18 in FIG. 4 . Here, too, the control unit 11 acquires the weight and urine volume measured during the treatment period for the symptom being treated. Note that the weight is stored as physical data in the electronic medical record DB 32a, and the urine volume is stored as test information in the electronic medical record DB 32a. Based on the acquired weight, the control unit 11 identifies the weight corresponding to each time unit of time indicated by the horizontal axis (the weight measured at each time or the weight assumed to have been measured at each time unit) and creates a graph (line graph) showing the change in weight over time for each unit of time. Furthermore, based on the acquired urine volume, the control unit 11 identifies the urine volume corresponding to each time unit of time indicated by the horizontal axis and creates a graph (bar graph) showing the change in urine volume over time for each unit of time. For example, when the unit of time indicated by the horizontal axis is one day as shown in FIG. 7 , the control unit 11 calculates the total daily urine volume for each time period or uses the value recorded as the daily urine volume.

[0049] The controller 11 also acquires the maximum and minimum diameters of the IVC and the respiratory variability of the selected patient from the electronic medical record database 32a. Here, the controller 11 also acquires the maximum and minimum diameters of the IVC and the respiratory variability measured during the treatment period for the symptom being treated. The maximum and minimum diameters of the IVC and the respiratory variability are stored as examination information in the electronic medical record database 32a. Based on the acquired data on the maximum and minimum diameters of the IVC, the controller 11 identifies the maximum and minimum diameters of the IVC corresponding to each unit of time indicated by the horizontal axis, and creates a graph (line graph) showing the temporal changes in the maximum and minimum diameters of the IVC for each unit of time. Based on the acquired respiratory variability of the IVC, the controller 11 also identifies the respiratory variability corresponding to each unit of time indicated by the horizontal axis, and creates a graph (bar graph) showing the temporal changes in the respiratory variability for each unit of time.

[0050] The control unit 11 then performs the process from step S19 onward and displays a detailed screen as shown in FIG. 7 on the display unit 15 (S23). Through this process, as shown in FIG. 7, in addition to the prescription details and edema rate at each date and time indicated on the horizontal axis, the weight, urine volume, maximum and minimum IVC diameters, and respiratory variation rate at each date and time can be displayed side by side. The weight and urine volume values ​​and their temporal changes can be used to more directly confirm the efficacy of the medication (diuretic) being administered. The maximum and minimum IVC diameters and the temporal changes in respiratory variation rate can be used to confirm the state and severity of heart failure symptoms as the therapeutic effect of the medication (diuretic) being administered. Therefore, by displaying this information side by side with the unit time indicated on the horizontal axis, the information that a physician should view when determining the medication prescription can be efficiently presented. This allows a physician to compare each item on the same time axis, which can assist in determining the medication prescription.

[0051] In the detailed screen shown in FIG. 7 , instead of or in addition to graphs of body weight and urine volume, region R3 may display graphs related to the patient's body composition, such as total body water, intracellular fluid volume, and extracellular fluid volume. By displaying this information, a physician can more quantitatively confirm the effects of the medication (diuretic) being administered. Furthermore, instead of or in addition to graphs of the IVC maximum and minimum diameters and respiratory variability, region R4 may display specialized test results, such as an edema score determined by palpation by a medical professional, a jugular venous distension score determined by visual inspection, test values ​​obtained using a thermodilution catheter (e.g., central venous pressure, pulmonary artery pressure, pulmonary artery wedge pressure, cardiac output, cardiac index), a cardiothoracic ratio determined by a chest X-ray, vascular biomarker test values ​​related to heart failure (e.g., BNP, NT-proBNP), and echocardiographic test values ​​(e.g., LVFE, LVDd, LAD, E / e′). In this case, too, by displaying each piece of information for each unit time indicated by the horizontal axis, various pieces of information can be easily compared along with the prescription contents and edema rate.

[0052] 7 screen, the line L indicating the date and time of the transition timing may not be added, and different backgrounds may be added to the area before the transition timing and the area after the transition timing, or only the line L indicating the date and time of the transition timing may be added. Alternatively, instead of the line L indicating the date and time of the transition timing, a mark (marker) may be added to the location of the date and time of the transition timing on the horizontal axis. The line L indicating the date and time of the transition timing may be added only to the edema rate graph, or may be added to at least one of the edema rate graph, the weight and urine volume graph, the maximum and minimum IVC diameters, and the respiratory variation rate graph. Furthermore, the different backgrounds added to the area before the transition timing and the area after the transition timing may be added only to the edema rate graph, or may be added to at least one of the edema rate graph, the weight and urine volume graph, the maximum and minimum IVC diameters, and the respiratory variation rate graph.

[0053] In this embodiment, as shown in Figures 6A to 7, the prescription contents and edema rate over time are displayed along with the unit time, and the patient's body composition values ​​or various test results are also displayed along with the unit time. Furthermore, in each graph, the transition timing when the drug administration method was switched from intravenous injection to oral administration is indicated by line L, allowing the user to grasp at a glance that the administration method was switched. Furthermore, in each graph, the area corresponding to the intravenous injection period and the area corresponding to the oral administration period are displayed with different backgrounds (different aspects), allowing the user to easily grasp each period and whether the patient is currently receiving medication via intravenous injection or oral administration. Because a patient's condition often worsens immediately after switching from intravenous injection to oral administration, physicians must pay attention to information regarding the period after the transition to oral administration. In the screens of Figures 6A to 7, the line L in the graph allows the user to easily grasp the timing of the transition, allowing the user to easily view information regarding the period of interest after the transition to oral administration. Therefore, information that can be used as a basis for determining whether a prescription change is necessary can be efficiently presented.

[0054] Each graph in the screens of Figures 6A to 7 may have any configuration as long as the region corresponding to the intravenous injection period and the region corresponding to the oral administration period can be easily identified, and is not limited to a configuration in which each region is displayed on a different background. For example, in each region, each plot point and each line segment of a line graph and each bar of a bar graph may be displayed in a different manner (e.g., different colors, different line types, different hatching). Furthermore, the shape of each graph in the screens of Figures 6A to 7 may be different from that shown in the drawings. For example, the line graphs displayed in the figures may be bar graphs, or they may be graphs showing cumulative values ​​or graphs showing cumulative values ​​and changes in value, such as waterfall charts. Furthermore, although Figures 6A to 7 are configured to be divided into regions R1 to R4, they may be divided into more regions or may be integrated into fewer regions (in extreme cases, a single region).

[0055] (Embodiment 2) In this embodiment, an information processing system is described that estimates a patient's future edema rate based on the patient's prescription information and time-series data of the edema rate and presents the estimated rate to a medical professional. The information processing system of this embodiment can be implemented using devices 10, 20, and 30 similar to those of the information processing system of embodiment 1 shown in FIGS. 1 and 2 , and therefore a description of the configurations of the devices 10, 20, and 30 will be omitted. In this embodiment, a learning model M that has learned training data through machine learning is stored in the memory unit 12 of the information processing device 10. The learning model M is expected to be used as a program module constituting artificial intelligence software. The learning model M performs a predetermined calculation on input values ​​and outputs the calculation results. Data such as coefficients and thresholds of functions that define this calculation are stored in the memory unit 12 as the learning model M. In addition to being configured to store the learning model M in the memory unit 12, the information processing device 10 may access a server that stores the learning model M and read it out.

[0056] Fig. 8 is an explanatory diagram showing an example of the configuration of the learning model M. The learning model M shown in Fig. 8 is trained to input the patient's prescription information, edema rate, and body weight, perform a calculation to predict the patient's future edema rate (e.g., the next day) based on the input information, and output the calculation result. The learning model M is configured using algorithms such as a convolutional neural network (CNN), a recurrent neural network (RNN), a long short-term memory (LSTM), a transformer, a decision tree, a random forest, and a support vector machine (SVM), and may be configured by combining multiple algorithms.

[0057] The learning model M has an input layer, an intermediate layer, and an output layer. The input layer has a plurality of input nodes, and each input node is associated with information to be input. The prescription information input to the learning model M includes all prescription information prescribed during the treatment period for the patient's symptoms, and for example, prescription details corresponding to each date and time (type of drug, dosage, administration method) are input via the input nodes. The edema rate and body weight can be time-series data, average values, or numerical values ​​indicating the trend of the edema rate (e.g., rate of change, moving average, etc.) of the edema rate and body weight measured during the treatment period.

[0058] The intermediate layer uses various functions, thresholds, etc. to calculate output values ​​from each piece of information input via the input layer and outputs the calculated output values ​​to the output layer. The output layer has multiple output nodes, each associated with a first edema rate, a second edema rate, etc., and outputs the probability (certainty) of estimating the associated edema rate. Each edema rate value associated with each output node can be any value within the range of edema rates that can be measured in a human body, for example, a value in the range of 0.20 to 0.50. The output value of each output node is, for example, a value between 0 and 1.0, and the sum of the probabilities output from each output node is 1.0 (100%). With the above-described configuration, when the patient's prescription details, edema rate, and weight information is input, the learning model M outputs information indicating the patient's future (e.g., the next day's) edema rate (certainty for each edema rate).

[0059] The information processing device 10 specifies, for example, the edema rate associated with the output node that outputs the largest output value (certainty) among the output values ​​from each output node in the above-described learning model M as the patient's future edema rate to be estimated. Note that the learning model M may be configured to have a single output node that outputs the edema rate with the highest confidence, instead of having multiple output nodes that output confidence levels for each edema rate.

[0060] The learning model M is generated by machine learning using training data including information on the prescription, edema rate, and weight of a training patient, and the patient's future (e.g., the next day) edema rate (correct label). For example, the training data is generated by assigning the patient's future (the next day's) edema rate to information including the prescription, edema rate, and weight for a patient who has developed edema symptoms due to heart failure. The training data generated in this manner is stored, for example, in a training DB (not shown) provided in the memory unit 12. Each piece of information used in the training data can be obtained from patient information stored in the electronic medical record DB 32a.

[0061] When the training data containing information on prescription details, edema rate, and body weight is input, the learning model M learns so that the output value from the output node corresponding to the edema rate 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 M performs calculations based on the input information and calculates output values ​​from each output node. The learning model M then compares the calculated output value of each output node with the value corresponding to the correct label (1 for the output node corresponding to the correct edema rate, and 0 for other output nodes) and optimizes parameters used in the calculation process so that the two values ​​approximate each other. These parameters are, for example, weights between neurons in the learning model M. The parameter optimization method is not particularly limited, and examples include backpropagation and steepest descent. This results in a learning model M that, when input with a patient's prescription details, edema rate, and body weight, is trained to estimate the patient's future edema rate and output the estimated result.

[0062] The learning model M may be trained by another learning device. The trained learning model M generated by training on another learning device is downloaded from the learning device to the information processing device 10 via, for example, the network N or a portable storage medium 10a and stored in the storage unit 12. The learning model M is not limited to the configuration shown in FIG. 8. The learning model M may be configured to receive various information used to estimate a patient's edema rate. For example, in addition to the patient's prescription, edema rate, and weight, the learning model M may also receive input of vital data such as urine volume (urine volume per unit time, total urine volume since the start of treatment, etc.), blood pressure, pulse pressure, body temperature, heart rate, pulse rate, and respiratory rate, body composition information such as extracellular fluid volume, total body water volume, and hydration rate, congestion index, medical history, primary cause of heart failure, treatment history, medication history, etc. The congestion index may be one or more of the following: the maximum diameter, minimum diameter, and respiratory variation rate of the IVC measured by echocardiography; test values ​​using a thermodilution catheter (central venous pressure, pulmonary artery pressure, pulmonary artery wedge pressure, cardiac output, cardiac index, etc.); edema score by palpation by a medical professional; jugular venous distension score by visual inspection; cardiothoracic ratio (cardiothoracic ratio) by chest X-ray; vascular biomarker test values ​​related to heart failure (BNP, NT-proBNP, etc.); and echocardiography values ​​(LVFE, LVDd, LAD, E / e', etc.). In this case, the learning model M is configured to estimate the edema rate for the next day based on various patient information, as well as the prescription content, edema rate, and weight. Furthermore, the learning model M is not limited to a configuration that estimates the edema rate for the next day, but may also be configured to estimate the edema rate for the future, such as two days, three days, or one week later. The learning model M may also be prepared for each attribute classified by age group and gender.

[0063] The following describes a process for estimating and presenting a patient's future (next day) edema rate using the learning model M while a doctor is checking the time-dependent change in the patient's edema rate using the screens shown in Figures 6A to 7. Figure 9 is a flowchart illustrating an example of the process for estimating a future edema rate, and Figures 10A and 10B are explanatory diagrams illustrating example screens. The control unit 11 of the information processing device 10 of this embodiment is capable of executing the same process as Figure 4. By executing the process of Figure 4, the display unit 15 displays a screen such as that shown in Figure 10A. The screen of Figure 10A displays a chart of prescription details and a graph of edema rate from the start of hospitalization through the fourth day, indicating that the medication is currently being administered via intravenous injection. The screen of Figure 10A also includes an edema rate estimation button for instructing execution of a process to estimate the patient's edema rate for the next day. Therefore, if a doctor wants to check the estimated edema rate for the next day, he or she operates the edema rate estimation button.

[0064] The control unit 11 determines whether the edema rate estimation button has been operated (S31). If it determines that the button has not been operated (S31: NO), the control unit 11 terminates the process. If it determines that the edema rate estimation button has been operated (S31: YES), the control unit 11 acquires, from the electronic medical record DB 32a of the electronic medical record server 30, each piece of information to be used for estimating the edema rate among the information of the patient being processed (S32). Specifically, when estimating the edema rate using the learning model M shown in FIG. 8, the control unit 11 acquires each piece of information on the patient's prescription, edema rate, and weight. Here, the control unit 11 acquires the prescription, edema rate, and weight within the treatment period.

[0065] Then, the control unit 11 estimates the future edema rate of the patient (specifically, the next time period relative to the horizontal axis setting, or the next day if the horizontal axis is in days) based on the acquired information (S33). Here, the control unit 11 inputs the acquired information on prescription details, edema rate, and weight into the learning model M, and estimates the future edema rate of the patient based on the output value from the learning model M. For example, the control unit 11 identifies the output node that outputs the largest output value (confidence factor) among the output values ​​from the learning model M, and identifies the edema rate associated with the identified output node as the future edema rate of the patient.

[0066] The control unit 11 adds the identified edema rate to the edema rate graph on the detailed screen (S34), as shown in FIG. 10B . At this time, the control unit 11 adds the plot points of the estimated edema rate to the graph in a manner different from the plot points currently being displayed. In the example of FIG. 10B , the plot points currently being displayed (plot points indicating measured edema rates) are shown as black circles, while the estimated edema rates are plotted as white circles. Furthermore, the lines connecting the plot points indicating measured edema rates are shown as solid lines, while the lines connecting the last measured edema rate and the estimated edema rate are shown as dashed lines. This configuration makes it easy to determine whether each plot point on the edema rate graph indicates a measured edema rate or an estimated edema rate.

[0067] This embodiment also achieves the same effects as the information processing system of the first embodiment. Furthermore, in this embodiment, on a screen displaying prescription details and edema rates side by side, a graph of the edema rate can be displayed with a future edema rate estimated from patient information. Therefore, a doctor can assess the patient's condition and determine whether to change the prescription, taking into account not only the patient's current medication status and edema rate, but also the estimated future edema rate. Furthermore, the present embodiment can also apply the modifications described in the first embodiment, as appropriate.

[0068] In this embodiment, the process of estimating a patient's future edema rate using the learning model M is not limited to being performed locally by the information processing device 10. For example, a server may be provided that executes the process of estimating a patient's future edema rate using the learning model M. In this case, the information processing device 10 can be configured to transmit information such as the patient's prescription information, edema rate, and weight to the server and receive the result (future edema rate) determined by the server.

[0069] (Embodiment 3) In this embodiment, an information processing system is described that estimates not only the future edema rate of a patient but also prescription changes to be made based on information such as the patient's prescription information and edema rate, and presents the results to a medical professional. The information processing system of this embodiment can be realized by devices 10, 20, and 30 similar to the information processing system of embodiment 1 shown in Figures 1 and 2, so a description of the configuration of each device 10, 20, and 30 will be omitted. Note that in this embodiment, the configuration of the learning model is different from that of the learning model M of embodiment 2.

[0070] Fig. 11 is an explanatory diagram showing an example of the configuration of a learning model Ma according to embodiment 3. The learning model Ma shown in Fig. 11 is trained to receive input of information such as a patient's prescription information, edema rate, weight, urine volume, body composition information, and congestion index, perform calculations to predict the patient's future (e.g., the next day's) edema rate and prescription information to be changed based on the input information, and output the calculation results. The learning model Ma may also be configured using algorithms such as CNN, RNN, LSTM, Transformer, decision tree, random forest, and SVM, or may be configured by combining multiple algorithms.

[0071] The prescription information, edema rate, and body weight input into the learning model Ma of this embodiment are the same as the prescription information, edema rate, and body weight input into the learning model M of embodiment 1. The urine volume input into the learning model Ma is the urine volume measured during the treatment period, and may be the urine volume per unit time (e.g., per day), the total urine volume during the treatment period, or the like. The body composition information may be time-series data, average values, or numerical values ​​(e.g., fluctuation rate, moving average, etc.) showing the progress of each body composition information, such as extracellular fluid volume, total body water volume, or hydration rate. The congestion index may be one or more of the maximum diameter, minimum diameter, and respiratory fluctuation rate of the IVC, test values ​​obtained using a thermodilution catheter, edema score obtained by palpation, jugular vein distension score obtained by visual inspection, cardiothoracic ratio obtained by chest X-ray, vascular biomarker test values ​​related to heart failure, echocardiography test values, etc.

[0072] Each output node in the output layer of the learning model Ma is associated with multiple edema rates and multiple pieces of prescription information, including the type of drug, dosage, and administration method (intravenous injection or oral administration). Each output node associated with an edema rate outputs a probability (certainty) that the corresponding edema rate should be estimated, and each output node associated with prescription information outputs a probability (certainty) that the corresponding prescription information should be estimated. The output value of each output node associated with an edema rate is, for example, a value between 0 and 1, the sum of which is 1 (100%), and the output value of each output node associated with prescription information is, for example, a value between 0 and 1, the sum of which is 1 (100%). With the above-described configuration, when various patient information is input, the learning model Ma outputs information indicating the patient's future edema rate (e.g., the next day) (certainty for each edema rate) and information indicating the prescription information to be changed (certainty for each prescription information).

[0073] The information processing device 10 specifies the edema rate of the output node that outputs the largest output value (certainty) among the output values ​​from the output nodes associated with the edema rate in the above-mentioned learning model Ma as the patient's future edema rate to be estimated. Furthermore, the information processing device 10 specifies the prescription information of the output node that outputs the largest output value (certainty) among the output values ​​from the output nodes associated with prescription information as the changed prescription information to be estimated. The learning model Ma may be configured to have one output node that outputs the edema rate with the highest confidence and one output node that outputs the prescription information with the highest confidence.

[0074] The learning model Ma is generated by machine learning using training data including various patient information for training and the patient's future (e.g., the next day) edema rate and prescription information (correct label). When each patient information included in the training data is input, the learning model Ma learns so that the output value from the output node corresponding to the correct edema rate approaches 1.0 and the output values ​​from the output nodes corresponding to other edema rates approach 0.0, and also so that the output value from the output node corresponding to the correct prescription information approaches 1.0 and the output values ​​from the output nodes corresponding to other prescription information approach 0.0. Again, during the learning process, the learning model Ma optimizes parameters such as the weights between neurons in the learning model Ma using backpropagation, steepest descent, or the like. This results in a learning model Ma that, when the above-mentioned patient information is input, estimates the patient's future edema rate and changed prescription information and outputs the estimation results.

[0075] The learning model Ma may also be learned by another learning device. Furthermore, the learning model Ma is not limited to the configuration shown in Fig. 11 , and may be configured, for example, similar to the learning model M shown in Fig. 8 , in which only the patient's prescription information, edema rate, and weight are input. Furthermore, the learning model Ma may be configured to input the patient's medical history, main cause of heart failure, type of heart failure, treatment history, readmission history, medication history, etc., in addition to the input data shown in Fig. 11 . Furthermore, the learning model Ma may be configured such that a model for estimating the edema rate and a model for estimating prescription information are separately provided.

[0076] The following describes a process for estimating and presenting the patient's future (next day) edema rate and changed prescription information using the learning model Ma. FIG. 12 is a flowchart showing an example of a process procedure for estimating a future edema rate according to the third embodiment. The process shown in FIG. 12 is the same as the process shown in FIG. 9 except that step S41 is added instead of step S33 and step S42 is added after step S34. Descriptions of steps that are the same as those in FIG. 9 will be omitted.

[0077] In step S32, the control unit 11 of the information processing device 10 of this embodiment acquires information used to estimate the edema rate and prescription information of the patient being processed from the electronic medical record DB 32a of the electronic medical record server 30. Specifically, the control unit 11 acquires information on the patient's prescription, edema rate, weight, urine volume, body composition information, and congestion index. Then, based on the acquired information, the control unit 11 estimates the patient's future (specifically, the next day's) edema rate and prescription information (S41). Here, the control unit 11 inputs the acquired information into a learning model Ma and estimates the patient's future edema rate and prescription information based on the output values ​​from the learning model Ma. For example, the control unit 11 identifies the output node that outputs the largest output value (confidence level) among the output values ​​from the output nodes associated with the edema rates of the learning model Ma, and identifies the edema rate associated with the identified output node as the patient's future edema rate. The control unit 11 also identifies the output node that output the largest output value (certainty) among the output values ​​from the output nodes to which the prescription information of the learning model Ma is associated, and identifies the prescription information associated with the identified output node as the prescription information to be changed for the patient.

[0078] The control unit 11 adds the identified edema rate to an edema rate graph on the details screen (S34), as shown in FIG. 10B , and displays the identified prescription information at a predetermined position on the details screen (S42). For example, the control unit 11 displays a message such as "It is estimated that the drug A2, dosage B2, can be changed to oral administration" at a predetermined position on the screen shown in FIG. 10B . This allows the healthcare professional to understand the estimated future change in edema rate and to receive a presentation of prescription changes that should be made, which the healthcare professional can use as advice when considering changing the prescription.

[0079] This embodiment also achieves the same effects as the information processing systems of the first and second embodiments. Furthermore, this embodiment can grasp the changed prescription content estimated from the patient's condition in addition to the future edema rate estimated from various patient information. Therefore, a doctor can determine the patient's condition by taking into account not only the patient's current medication status and edema rate, but also the estimated future edema rate and changed prescription information, and can decide whether or not to change the prescription content. Furthermore, the modified examples described in the first and second embodiments can also be applied to this embodiment.

[0080] In this embodiment, too, the process of estimating the patient's future edema rate and changed prescription information using the learning model Ma is not limited to being performed locally by the information processing device 10. For example, a server may be provided that executes the process of estimating the patient's future edema rate and changed prescription information using the learning model Ma. In this case, the information processing device 10 can be configured to transmit each piece of information on the patient's prescription information, edema rate, weight, urine volume, body composition information, and congestion index to the server and receive the results determined by the server (the future edema rate and changed prescription information).

[0081] The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. 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.

[0082] In each of the above-described embodiments, the index showing the degree of edema is not limited to the edema index, but may also be a value obtained by dividing the extracellular water content by the intracellular water content, a total body water percentage which is the ratio of total body water to body weight, an evaluation value on an edema assessment scale, an estimated congestion value estimated from image data of a patient, etc.

[0083] The features described in each of the above-mentioned embodiments can be combined with each other. Furthermore, the independent claims and dependent claims described in the claims can be combined with each other in any and all combinations, regardless of the reference format. Furthermore, while the claims use a format in which a claim references two or more other claims (multiple claim format), this is not limited to this format. A multiple claim (multi-multi claim) that references at least one other multiple claim may also be used.

[0084] REFERENCE SIGNS LIST 10 Information processing device 11 Control unit 12 Memory unit 13 Communication unit 14 Input unit 15 Display unit 20 Body composition monitor 30 Electronic medical record server 31 Control unit 32 Memory unit 33 Communication unit 34 Input unit 35 Display unit 32a Electronic medical record DB M Learning model Ma Learning model

Claims

1. A program that causes a computer to execute the following processes: storing prescription information of medications prescribed to a subject in a storage unit; storing time-series data of an index indicating the degree of edema of the subject in a storage unit; and outputting a first screen that displays, in association with date and time, information on the medications administered at each date and time and the index indicating the degree of edema measured at each date and time.

2. The program according to claim 1, which causes the computer to execute the following process: based on the time series data of the index indicating the degree of edema, create a graph showing the time series change of the index indicating the degree of edema, with the first axis showing the date and time and the second axis showing the value of the index; and generate the first screen which displays the graph alongside information on the medication administered at each date and time, corresponding to each date and time shown on the first axis of the graph.

3. The program according to claim 2, which causes the computer to execute a process of adding a mark indicating a normal range set for the index indicating the degree of edema onto the graph in the first screen.

4. The program according to claim 2 or 3, which causes the computer to execute the following process: based on the prescription information or based on date and time specification information received from the user, specify the date and time when the administration method of the drug was changed from intravenous injection to oral administration; add a line indicating the specified date and time to the drug information and the graph in the first screen; and add different backgrounds to the area before the date and time indicated by the line and the area after the date and time indicated by the line.

5. The program according to claim 2 or 3, which causes the computer to execute the process of identifying the date and time when the administration method of the drug was changed from intravenous injection to oral administration based on the prescription information or based on date and time identification information received from the user, and adding a line or mark indicating the identified date and time on the graph in the first screen.

6. The program according to claim 2 or 3, which causes the computer to execute the process of identifying the date and time when the administration method of the drug was changed from intravenous injection to oral administration based on the prescription information or based on date and time identification information received from the user, and adding different backgrounds to the area before the identified date and time and the area after the identified date and time on the graph in the first screen.

7. The program according to claim 2 or 3, which causes the computer to execute the process of: calculating the dosage of the drug administered at each time interval based on the prescription information and the time interval indicating the measurement interval of the index indicating the degree of edema; and generating the first screen which displays information on the drug administered at each time interval in association with the first axis of the graph.

8. The program according to claim 2 or 3, which causes the computer to execute the following process: inputting the prescription information of the subject and the time series data of an index indicating the degree of edema stored in the memory unit, and the subject's physical data, into a learning model that has been trained to output an index indicating the subject's future degree of edema when the prescription information of the subject, time series data of an index indicating the degree of edema, and the subject's physical data are input, to obtain an index indicating the subject's future degree of edema; and adding the obtained index indicating the future degree of edema to the graph on the first screen, in association with a future date and time.

9. The program according to claim 8, which causes the computer to execute a process of adding the index indicating the degree of future edema in a manner different from other values ​​of the index in the graph.

10. A program as claimed in any one of claims 1 to 3, which causes the computer to execute the following processes: outputting a second screen displaying prescription information for a plurality of subjects, including the type of prescribed medication, dosage, and administration method, as well as the number of days since the administration method was changed from intravenous injection to oral administration, in association with an index indicating the degree of edema and / or the trend of said index over time; accepting the selection of one of the plurality of subjects via said second screen; and outputting the first screen for the selected subject.

11. The program according to any one of claims 1 to 3, wherein the index indicating the degree of edema is an edema index calculated as the ratio of extracellular fluid volume (ECW) to total body water (TBW).

12. An information processing method in which a computer executes the following processes: storing prescription information of medication prescribed to a subject in a storage unit; storing time series data of an index indicating the degree of edema of the subject in a storage unit; and outputting a first screen that displays, in association with date and time, information on the medication administered at each date and time and the index indicating the degree of edema measured at each date and time.

13. An information processing device having a control unit, wherein the control unit stores prescription information of medication prescribed to a subject in a memory unit, stores time series data of an index indicating the degree of edema of the subject in the memory unit, and outputs a first screen that displays information on the medication administered at each date and time and an index indicating the degree of edema measured at each date and time, in association with the date and time.

Images