Program, information processing method, and information processing apparatus

Abstract

Provided are a program and the like for displaying an index suitable for comprehending the condition of a kidney.　The program causes a computer to execute a process for: acquiring urinary information including a urinary oxygen partial pressure; calculating a renal state index on the basis of the urinary information; and outputting the renal state index. The urinary information includes, in addition to the urinary oxygen partial pressure, a urine volume, a urinary flow rate, a urine color, the absorbance of urine, a urinary sodium amount, or a urinary creatinine amount. The renal state index is an index pertaining to an estimate of the renal artery blood flow rate. The urinary information includes the urinary oxygen partial pressure and the urinary flow rate. The renal state index is calculated on the basis of the urinary oxygen partial pressure and the urinary flow rate, by using a predetermined algorithm.

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] A catheter for measuring the partial pressure of oxygen in urine in the bladder has been proposed, which combines a urethral catheter body with a probe that is inserted into the bladder via the urethral catheter to measure the partial pressure of oxygen in urine in the bladder (Patent Document 1).

[0003] Japanese Patent Application Laid-Open No. 2021-62073

[0004] The urinary oxygen partial pressure measured by the catheter in Patent Document 1 is one of the parameters indicating the state of the kidneys. Medical professionals such as doctors and nurses comprehensively assess the urinary oxygen partial pressure and vital data such as blood pressure to understand the state of the kidneys of a patient.

[0005] However, for example, during surgery and during systemic management in an ICU (Intensive Care Unit), medical personnel are performing various tasks simultaneously and are not able to concentrate on understanding the state of the kidneys.

[0006] In one aspect, an object of the present invention is to provide a program or the like that displays indicators suitable for understanding the state of the kidneys.

[0007] The program causes the computer to execute processes of acquiring urine information including urinary oxygen partial pressure, calculating a renal condition index based on the urine information, and outputting the renal condition index.

[0008] In one aspect, it is possible to provide a program or the like that displays indicators suitable for understanding the state of the kidneys.

[0009] FIG. 1 is an explanatory diagram illustrating the configuration of an information processing system. FIG. 2 is a flowchart illustrating the processing flow of a program. FIG. 3 is an example screen. FIG. 4 is an example screen. FIG. 5 is an example screen. FIG. 6 is an explanatory diagram illustrating a renal status model. FIG. 7 is an explanatory diagram illustrating a treatment guideline model. FIG. 8 is a flowchart illustrating the processing flow of a program of embodiment 3. FIG. 9 is an example screen of embodiment 3. FIG. 10 is an explanatory diagram illustrating a renal status model of embodiment 4. FIG. 11 is an example screen of embodiment 4. FIG. 12 is an explanatory diagram illustrating a drug model. FIG. 13 is a flowchart illustrating the processing flow of a program of embodiment 5. FIG. 14 is an explanatory diagram illustrating the configuration of an information processing system of embodiment 6.

[0010] [Embodiment 1] It is known that patients who require surgery and whole-body management in an ICU are at high risk of developing acute kidney injury (AKI), a condition in which the patient's renal function declines in a short period of time. Because the prognosis of AKI is poor, it is desirable to constantly monitor the state of the patient's kidneys, detect signs of AKI, and take appropriate preventive measures before AKI actually occurs.

[0011] Renal status is evaluated based on evaluation indices such as serum creatinine level, creatinine clearance, glomerular filtration rate (GFR), and urine flow rate. However, serum creatinine level changes approximately one to three days after a kidney injury event. Since creatinine clearance and glomerular filtration rate are calculated using serum creatinine level, they change one to three days after kidney injury, similar to serum creatinine level. It is known that urine flow rate may not show any change even when kidney injury occurs.

[0012] Therefore, it is difficult to quickly detect signs of acute kidney injury and take preventive measures using conventional evaluation indices alone.

[0013] It is known that changes in renal tissue oxygen tension quickly reflect changes in the state of the kidney. However, measuring renal tissue oxygen tension requires inserting an oxygen sensor into the renal tissue. Inserting an oxygen sensor into renal tissue is highly invasive to the patient and can damage the kidney.

[0014] On the other hand, during surgery and during systemic management in the ICU, an indwelling bladder catheter is often used. Therefore, the catheter disclosed in Patent Document 1, for example, can be used to measure the urinary oxygen partial pressure in the bladder in real time without increasing the degree of invasiveness to the patient.

[0015] The inventors of the present invention measured the changes in renal tissue oxygen tension, urinary oxygen tension, urine flow rate, and pulse pressure while changing the concentration of anesthesia respiratory gas in pigs. Pulse pressure is the difference between systolic and diastolic blood pressure. The sample size, i.e., the number of pigs used, was six.

[0016] Based on the measurement data, the correlation coefficient r between urinary oxygen partial pressure and renal tissue oxygen partial pressure was calculated to be 0.6581. The calculated correlation coefficient r was equivalent to the value in the prior art, confirming that the measurements were performed appropriately.

[0017] Using the measurement data obtained in this experiment, three types of multiple regression analysis were performed, and the results are shown in Table 1.

[0018]

[0019] The renal medulla is a type of renal tissue located inside the kidney. Therefore, renal medullary oxygen partial pressure is an example of renal tissue oxygen partial pressure. Table 1 shows that the renal medullary oxygen partial pressure, which is the objective variable, can be estimated with high accuracy by using a multiple regression equation calculated using three data points: urinary oxygen partial pressure, urine flow rate, and pulse pressure, as explanatory variables.

[0020] Furthermore, Table 1 shows that by using a multiple regression equation that uses urinary oxygen partial pressure and urine flow rate as explanatory variables, the target variable, renal medullary oxygen partial pressure, can be estimated with high accuracy, at least as accurately as a multiple regression equation that uses urinary oxygen partial pressure and pulse pressure as explanatory variables.

[0021] The calculation method for each explanatory variable will be explained below. As mentioned above, the urinary oxygen partial pressure can be measured in real time by inserting an oxygen sensor into the bladder via an indwelling bladder catheter. The oxygen sensor may be placed midway through the indwelling bladder catheter, or midway between the end of the indwelling bladder catheter and the entrance of a urine bag connected to the end of the catheter by a tube.

[0022] The urine flow rate can be measured in real time, for example, based on changes in the weight of a urine bag. A flow rate sensor may be attached to an indwelling bladder catheter. The type of sensor used to measure urine flow rate is not important. Patients undergoing surgery or undergoing general care in the ICU are usually monitored constantly for circulatory dynamics information, including blood pressure and heart rate, using a vital sign monitor 33 (see Figure 1).

[0023] As described above, by using a multiple regression equation, it is possible to estimate renal medullary oxygen partial pressure with high accuracy in real time without increasing the degree of invasiveness to the patient.

[0024] Renal medullary oxygen partial pressure varies greatly from person to person depending on the patient's hemodynamics and management status. Therefore, signs of changes in renal medullary oxygen partial pressure that could lead to acute kidney injury should be determined based not only on the absolute value of renal medullary oxygen partial pressure, but also on changes from the initial state, such as at the start of surgery or admission to the ICU.

[0025] In this embodiment, an index relating to the estimated value of renal medullary oxygen partial pressure is displayed as a relative value, with the initial value at the start of surgery or admission to the ICU being set to 100. That is, the displayed value is calculated using equation (1).

[0026]

[0027] Urinary oxygen partial pressure and urine flow rate are examples of urine information of a patient. The urine information may be, for example, urine volume, urine color, urine absorbance, urine sodium content, urine potassium content, or urine creatinine content. Urine volume is the total amount of urine excreted from the patient's bladder after measurement begins, and can be measured in real time based on the weight of the urine bag, etc.

[0028] Urine color and absorbance can be measured in real time by inserting an optical fiber connected to an optical measuring device such as a spectrophotometer into an indwelling bladder catheter. Urinary creatinine levels can be measured using an enzymatic method or absorbance. Urinary sodium and potassium levels can also be measured in real time by using fluorescent dyes that react specifically within the system from the indwelling bladder catheter to the urine bag.

[0029] Pulse pressure is an example of circulatory dynamics information. Circulatory dynamics information may be, for example, systolic blood pressure, diastolic blood pressure, mean blood pressure, heart rate, or blood oxygen saturation. Mean blood pressure is calculated by diastolic blood pressure + (systolic blood pressure - diastolic blood pressure) / 3. However, numerical processing such as moving averages and adjustments to the calculation formula may be performed during calculation, as long as equivalent values ​​are used.

[0030] These pieces of circulatory dynamics information can be measured by the vital signs monitor 33. The circulatory dynamics information may be tissue oxygen saturation in various parts of the body. This monitor can measure tissue oxygen saturation in the brain or various parts of the body in real time with minimal invasiveness, and is used in clinical settings as a hemodynamics monitor.

[0031] Renal medullary oxygen partial pressure is an example of a renal condition index that can quickly detect changes in the condition of the kidney. The renal condition index may be, for example, an estimated value of renal artery blood flow. When using an estimated value of renal artery blood flow, data that allows renal artery blood flow to be used as the objective variable in the above-mentioned experiment is obtained and multiple regression analysis is performed. In addition, any index that can detect kidney abnormalities can be used.

[0032] The multiple regression equation calculated using urinary oxygen partial pressure and urine flow rate as explanatory variables and renal medullary oxygen partial pressure as a response variable, as explained in No. 1 of Table 1, is an example of an algorithm for calculating a renal condition index based on urinary oxygen partial pressure and urine flow rate.

[0033] In the aforementioned measurements using pigs, in addition to urinary oxygen partial pressure, urine flow rate, and renal medullary oxygen partial pressure, mean blood pressure, pulse pressure, heart rate, renal artery blood flow, and glomerular filtration rate can also be measured. A multiple regression equation can be calculated using urinary oxygen partial pressure, urine flow rate, mean blood pressure, pulse pressure, and heart rate as explanatory variables and renal artery blood flow rate as the dependent variable. The multiple regression equation calculated in this way is an example of an algorithm for calculating a renal condition index related to an estimated value of renal artery blood flow based on urinary oxygen partial pressure, urine flow rate, mean blood pressure, pulse pressure, and heart rate.

[0034] Using the same experimental data, a multiple regression equation can be calculated using urinary oxygen partial pressure, urine flow rate, mean blood pressure, pulse pressure, and heart rate as explanatory variables and renal medullary oxygen partial pressure as a response variable. The multiple regression equation calculated in this way is an example of an algorithm for calculating a renal condition index related to an estimated value of renal tissue oxygen partial pressure based on urinary oxygen partial pressure, urine flow rate, mean blood pressure, pulse pressure, and heart rate.

[0035] A multiple regression equation can be calculated using urinary oxygen partial pressure and urine flow rate as explanatory variables and renal artery blood flow rate as a response variable. The multiple regression equation calculated in this way is an example of an algorithm for calculating a renal condition index related to an estimated value of renal artery blood flow rate based on urinary oxygen partial pressure and urine flow rate.

[0036] A simple regression equation may be used instead of a multiple regression equation. For example, a simple regression equation can be calculated using urinary oxygen partial pressure or urine flow rate as an explanatory variable and renal artery blood flow rate as a response variable. The simple regression equation calculated in this manner is an example of an algorithm for calculating a renal condition index related to an estimated value of renal artery blood flow rate based on either urinary oxygen partial pressure or urine flow rate.

[0037] A multiple regression equation or a simple regression equation may be calculated using the glomerular filtration rate instead of the renal artery blood flow as an explanatory variable. The multiple regression equation or the simple regression equation calculated in this manner is an example of an algorithm for calculating a renal condition index related to an estimated value of the glomerular filtration rate.

[0038] Additionally, it is possible to generate an algorithm for calculating any index by calculating a multiple regression equation using any combination of measured parameters.

[0039] A table showing the relationship between explanatory variables and response variables may be created based on the multiple regression equation and stored in the auxiliary storage device 23 (see FIG. 1). When there are two explanatory variables, urinary oxygen partial pressure and urine flow rate, and the response variable is renal artery blood flow, a three-dimensional table is created. When there are five explanatory variables, urinary oxygen partial pressure, urine flow rate, mean blood pressure, pulse pressure, and heart rate, and the response variable is renal medullary oxygen partial pressure, a six-dimensional table is created.

[0040] The control unit 21 (see FIG. 1) searches the table using explanatory variables measured in real time as keys to obtain the objective variable. The control unit 21 may calculate the objective variable by interpolating between parameters recorded in the table using a known method. The table is an example of an algorithm for calculating the objective variable based on the explanatory variables.

[0041] In the following description, the renal condition index is renal medullary oxygen partial pressure, and an example will be described in which renal medullary oxygen partial pressure is estimated based on urinary oxygen partial pressure, urine flow rate, and pulse pressure.

[0042] 1 is an explanatory diagram illustrating the configuration of an information processing system 10. The information processing system 10 includes an information processing device 20, a urine measurement device 31, a vital sign monitor 33, a syringe pump 34, an infusion pump 35, and an electronic medical record system 17, which are connected via a network such as an HIS (Hospital Information System).

[0043] The urine measurement device 31 measures and outputs urine information such as urinary oxygen partial pressure and urine flow rate using a urine sensor 311. The vital monitor 33 is connected to various sensors (not shown) and measures and outputs circulatory dynamics information such as blood pressure and pulse pressure. The urine measurement device 31 and the vital monitor 33 may be configured integrally.

[0044] The syringe pump 34 is used to administer medicine to a patient. The infusion pump 35 is used to administer an infusion to a patient. The infusion pump 35 may have a function of mixing medicine with an infusion. Instead of using the syringe pump 34, a nurse or the like may administer medicine to a patient using a syringe. The infusion pump 35 may not be connected to a network and may be manually operated by a nurse or the like. The syringe pump 34 and the infusion pump 35 are examples of medicine administration devices.

[0045] The electronic medical record system 17 records patient information such as the patient's age, sex, height, weight, biochemical test results, medical history, and ongoing treatment progress. The ongoing treatment progress includes the type and amount of medication administered to the patient. The medical history includes the type and amount of medication administered to the patient in past surgeries and treatments, and the patient's condition after administration. Data such as renal condition indicators, which will be described later, may be sequentially recorded in the electronic medical record system 17.

[0046] The information processing device 20 includes a control unit 21, a main memory device 22, an auxiliary memory device 23, a communication unit 24, a display unit 25, an input unit 26, and a bus. The control unit 21 is an arithmetic and control device that executes the program of this embodiment. The control unit 21 uses one or more central processing units (CPUs), graphics processing units (GPUs), tensor processing units (TPUs), multi-core CPUs, or the like. The control unit 21 is connected to each hardware unit that constitutes the information processing device 20 via the bus.

[0047] The main memory device 22 is a storage device such as an SRAM (Static Random Access Memory), a DRAM (Dynamic Random Access Memory), a flash memory, etc. The main memory device 22 temporarily stores information required during processing performed by the control unit 21 and programs currently being executed by the control unit 21.

[0048] The auxiliary storage device 23 is a storage device such as an SRAM, a flash memory, a hard disk, or a magnetic tape. The auxiliary storage device 23 stores programs to be executed by the control unit 21 and various data required for executing the programs. The communication unit 24 is an interface for communication between the information processing device 20 and a network.

[0049] The display unit 25 is, for example, a liquid crystal display device or an organic EL (Electro Luminescence) display device. The input unit 26 is, for example, an input device such as a keyboard, a mouse, a trackball, or a microphone. The display unit 25 and the input unit 26 may be stacked together to form a touch panel.

[0050] The information processing device 20 in this embodiment is an information device such as a general-purpose personal computer, a tablet, a smartphone, or a server computer. The information processing device 20 may be a mainframe computer, a virtual machine running on a mainframe computer, a cloud computing system, a quantum computer, or multiple personal computers performing distributed processing. The information processing device 20 may be integrated with, for example, the urine measurement device 31, the vital sign monitor 33, or the electronic medical record system 17.

[0051] The following description will be made mainly on the assumption that the control unit 21 performs software processing. The processing and various models described using flowcharts may be realized by dedicated hardware.

[0052] 2 is a flowchart illustrating the flow of processing of the program. After attaching various sensors to the body of a patient who has entered an operating room or an ICU, a user such as a doctor starts the program described using FIG. 2.

[0053] The control unit 21 acquires urine information from the urine measurement device 31 (step S501). The control unit 21 acquires circulatory dynamics information from the vital sign monitor 33 (step S502). The control unit 21 calculates a renal status index by substituting the urine information acquired in step S501 and the circulatory dynamics information acquired in step S502 into a multiple regression equation calculated in advance (step S503). The renal status index calculated in step S503 is the initial value of the renal status index.

[0054] The control unit 21 acquires urine information from the urine measurement device 31 (step S504). The control unit 21 acquires circulatory dynamics information from the vital sign monitor 33 (step S505). The control unit 21 calculates a renal status index by substituting the urine information acquired in step S504 and the circulatory dynamics information acquired in step S505 into a multiple regression equation calculated in advance (step S506). The renal status index calculated in step S506 is a real-time renal status index.

[0055] The control unit 21 calculates a dimensionless renal status index by dividing the real-time renal status index acquired in step S506 by the initial value of the renal status index acquired in step S503 and multiplying the result by 100 (step S507). The dimensionless index at the start of measurement is 100. Note that if the real-time renal status index exceeds the initial value, the dimensionless index will be a value exceeding 100.

[0056] The control unit 21 displays the calculated renal status index on the display unit 25 (step S508). The control unit 21 may output the renal status index to another device connected to the output device, such as the vital sign monitor 33 or the electronic medical record system 17.

[0057] The control unit 21 determines whether to terminate the process (step S509). For example, if an instruction to terminate is received from the user, the control unit 21 determines to terminate the process. The control unit 21 may also determine to terminate the process if it becomes unable to acquire urine information and circulatory dynamics information from the urine measurement device 31 and the vital sign monitor 33.

[0058] If it is determined not to end the process (NO in step S509), the control unit 21 returns to step S504. If it is determined to end the process (YES in step S509), the control unit 21 ends the process.

[0059] Figures 3 to 6 are example screens. The control unit 21 displays the screens shown in Figures 3 to 6 in step S508 of the program described using Figure 2. The control unit 21 may display the screens shown in Figures 3 to 6 on a display device located in, for example, a nurse's center. The control unit 21 may also receive a selection from the user as to which screen to display.

[0060] The screen shown in Figure 3 has a renal status index field 72, a connection status field 71, a urinary oxygen partial pressure field 731, a urine flow field 732, a mean blood pressure field 733, a heart rate field 734, and a patient information field 74. A menu button is located in the upper left corner of the screen, and an indicator showing the battery charge state is also located in the upper left corner of the screen. To the right of the menu button, the date, day of the week, time, temperature, and air pressure are displayed. The temperature and air pressure are obtained, for example, from a sensor module connected to the information processing device 20 that can obtain temperature and air pressure. They may also be obtained from a World Wide Web (WWW) site that provides weather information via a network.

[0061] The renal condition index field 72 is located in the center of the screen, and displays the dimensionless index calculated in step S507 in real time. A renal condition indicator 721 is located next to the renal condition index field 72. The renal condition indicator 721 has an inverted triangular shape, and illustrates the renal condition index field 72 in three stages: less than 30 percent, 30 to 70 percent, and 70 percent or more.

[0062] The urinary oxygen partial pressure field 731 displays the urinary oxygen partial pressure from the urine information acquired in step S501. The urine flow rate field 732 displays the urine flow rate from the urine information acquired in step S501. The mean blood pressure field 733 displays the mean blood pressure from the circulatory dynamics information acquired in step S502. The heart rate field 734 displays the heart rate from the circulatory dynamics information acquired in step S502.

[0063] The patient information field 74 displays patient information obtained from the electronic medical record system 17. In the example shown in Fig. 3, the patient weighs 54 kilograms, is 167 centimeters tall, is male, and is 59 years old.

[0064] 3, an icon representing the syringe pump 34 and an icon representing the infusion pump 35 are displayed in the connection status field 71. The user can confirm that the syringe pump 34 and the infusion pump 35 are both operational and ready for use. For example, if multiple syringe pumps 34 are connected to the network, the control unit 21 may display only one icon representing the syringe pump 34, or may display as many icons as the number of connected syringe pumps 34.

[0065] In the screen shown in Figure 4, a renal status graph 75 is arranged to the right of the renal status index field 72. Everything except the renal status graph 75 is the same as in Figure 3, so a description thereof will be omitted. The horizontal axis of the renal status graph 75 represents time. The right end of the horizontal axis represents the current time. The vertical axis of the renal status graph 75 represents the non-dimensional renal status index. The renal status graph 75 allows the user to easily check the change over time in the dimensionless index.

[0066] The renal status graph 75 is displayed as a line graph extending to the right after measurement begins. As shown in Figure 4, when the right end of the renal status graph 75 is reached, the controller 21 scrolls the renal status graph 75 to the right by one piece of data each time new data is acquired, and displays the most recent data at the right end of the renal status graph 75. Alternatively, for example, all data may be displayed in a compressed time format.

[0067] 5, a range indicator 751 indicating the range of variation of the renal status index is displayed at the left end of the renal status graph 75. The upper and lower ends of the thick part of the range indicator 751 correspond to the maximum and minimum values, respectively, of the renal status index displayed on the renal status graph 75. Error bars indicating the calculation error of the renal status index are displayed above and below the range indicator 751. The range indicator 751 allows the user to easily understand the range of variation and error range of the renal status index.

[0068] 6, a urinary oxygen partial pressure column 731, a urine flow rate column 732, a mean blood pressure column 733, a heart rate column 734, an arterial oxygen saturation column 735, and an inhaled oxygen concentration column 736 are arranged in a tiled pattern. The urinary oxygen partial pressure column 731 displays a graph showing the change in urinary oxygen partial pressure over time and the latest urinary oxygen partial pressure. The horizontal axis of the graph in the urinary oxygen partial pressure column 731 represents time, and the vertical axis represents urinary oxygen partial pressure.

[0069] The urine flow field 732 displays a graph showing changes in urine flow over time and the latest urine flow. The horizontal axis of the graph in the urine flow field 732 represents time, and the vertical axis represents urine flow. The mean blood pressure field 733 displays a graph showing changes in mean blood pressure over time and the latest mean blood pressure. The horizontal axis of the graph in the mean blood pressure field 733 represents time, and the vertical axis represents mean blood pressure. The heart rate field 734 displays a graph showing changes in heart rate over time and the latest heart rate. The horizontal axis of the graph in the heart rate field 734 represents time, and the vertical axis represents heart rate.

[0070] A graph showing changes in arterial blood oxygen saturation (SpO2: Saturation of percutaneous oxygen) over time and the latest arterial blood oxygen saturation are displayed in the arterial blood oxygen saturation field 735. The horizontal axis of the graph in the arterial blood oxygen saturation field 735 represents time, and the vertical axis represents arterial blood oxygen saturation.

[0071] The fraction of inspiratory oxygen field 736 displays a graph showing the change in fraction of inspiratory oxygen (FiO2) over time and the latest fraction of inspiratory oxygen. The horizontal axis of the graph in the fraction of inspiratory oxygen field 736 represents time, and the vertical axis represents the fraction of inspiratory oxygen. The fraction of inspiratory oxygen is calculated from the type of oxygen administration device attached to the patient, such as a nasal cannula, a mask with a reservoir, or a Venturi mask, and the oxygen flow rate. If the patient is under general anesthesia, the fraction of inspiratory oxygen may be obtained from an anesthesia machine (not shown).

[0072] The horizontal axes of the six graphs shown in Fig. 6 are all the same. Therefore, instead of arranging the six graphs in a tiled pattern, six line graphs may be displayed on one graph. The graphs shown in Figs. 4 to 6 may also be bar graphs.

[0073] The screen examples described using Figures 3 to 6 are all illustrative. The control unit 21 may accept an instruction from the user to change the items to be displayed on the screen and the layout of each item.

[0074] 7 is an example screen. In FIG. 7, a condition selection field 761 is displayed on the left side of the screen, and a correlation graph field 762 is displayed on the right side of the screen. The horizontal axis of the graph shown in the correlation graph field 762 is the item selected in the "X-axis item" of the condition selection field 761. The vertical axis of the graph shown in the correlation graph field 762 is the item selected in the "Y-axis item" of the condition selection field 761.

[0075] The control unit 21 plots the data averaged over the time period selected in the "data averaging interval" in the correlation graph field 762. The items selected in the "calculation parameters" are displayed at the top of the correlation graph field 762. In FIG. 7, an approximation formula that approximates the plotted data with a linear function and the Pearson correlation coefficient are displayed. The user can use the screen in FIG. 7 to check the relationship between data items.

[0076] According to this embodiment, an information processing system 10 can be provided that outputs a renal status index suitable for understanding the state of the kidneys. By displaying dimensionless renal status, it is possible to assist the user in early detection of signs of acute kidney injury and taking necessary measures. Because it is not necessary to simultaneously grasp both the urinary oxygen partial pressure and urine flow rate measured by the urine measurement device 31 and the pulse pressure measured by the vital sign monitor 33, it is possible to provide an information processing system 10 that reduces the burden on the user. Note that urinary oxygen concentration may be used instead of urinary oxygen partial pressure.

[0077] According to this embodiment, an information processing system 10 can be provided that allows the user to easily grasp the change in the renal status index over time using the renal status graph 75 shown in Figures 4 and 5. For example, if the renal status index suddenly decreases, it can be determined that there is a high risk of developing acute kidney injury. Note that if the value of the renal status index or the slope of the renal status index exceeds a predetermined threshold, the control unit 21 may alert the user by sounding an alarm or flashing the display unit 25, for example.

[0078] According to this embodiment, an information processing system 10 can be provided that allows a user to simultaneously grasp multiple items using the screen described using Fig. 6. Note that the combination of graphs shown in Fig. 6 is an example. It is desirable that the user be able to select the combination of graphs to be displayed. For example, the renal status graph 75 shown in Figs. 4 and 5 may be one of the graphs displayed on the screen of Fig. 6.

[0079] According to this embodiment, it is possible to provide an information processing system 10 that allows the user to grasp the interrelationships between a plurality of data items using the screen described with reference to FIG.

[0080] [Embodiment 2] This embodiment relates to an information processing system 10 that uses a trained model generated by machine learning instead of using a multiple regression equation. Description of parts common to embodiment 1 will be omitted.

[0081] 8 is an explanatory diagram illustrating the renal condition model 42. The renal condition model 42 is a trained model that receives input of explanatory variables and outputs a target variable. The present inventors used the experimental data described in the first embodiment as training data and generated two renal condition models 42 through machine learning using the XGBoost algorithm. The explanatory variables, target variables, and coefficients of determination of the renal condition model 42 are shown in Table 2.

[0082]

[0083] Both No. 4 and No. 5 have larger coefficients of determination than the multiple regression equation shown in Table 1, which indicates that they can estimate renal medullary oxygen partial pressure with greater accuracy than the multiple regression equation. From No. 1 shown in Table 1 and No. 4 shown in Table 2, it can be seen that by using renal state model 42, renal medullary oxygen partial pressure can be estimated with greater accuracy based on two parameters, urinary oxygen partial pressure and urine flow rate.

[0084] XGBoost is an example of a machine learning algorithm. The trained model may be generated using an algorithm such as Random Forest or Convolutional Neural Network (CNN). The renal status model 42 shown in Table 2 is an example.

[0085] A renal condition model 42 may be generated using explanatory variables and response variables different from those in No. 4 and No. 5. For example, patient information such as the patient's age, sex, height, weight, biochemical test results, medical history, and ongoing treatment progress may be included in the explanatory variables. By adding patient information to the explanatory variables, a renal condition model 42 with a highly accurate output response variable can be generated.

[0086] The renal state model 42 is stored in the auxiliary storage device 23 or an external large-capacity storage device connected to the information processing device 20. In this embodiment, in steps S503 and S506 of the flowchart described with reference to Fig. 2, the control unit 21 inputs explanatory variables to the renal state model 42 and acquires the renal medullary oxygen partial pressure.

[0087] The control unit 21 may select and use, for example, the No. 4 renal condition model 42. The control unit 21 may calculate the objective variable and reliability using each of the multiple renal condition models 42, and then adopt the objective variable output by the renal condition model 42 with the highest reliability. The control unit 21 may adopt a representative value of the multiple objective variables calculated using each of the multiple renal condition models 42. The representative value may be any statistical value, such as an arithmetic mean value, a geometric mean value, a maximum value, or a minimum value.

[0088] According to the present embodiment, it is possible to provide an information processing system 10 that displays a renal status index with high accuracy. By using a highly accurate renal status index, it is possible to provide an information processing system 10 that displays highly reliable information regarding the state of the kidneys.

[0089] Third Embodiment This embodiment relates to an information processing system 10 that displays treatment guidelines. Explanation of parts common to the second embodiment will be omitted.

[0090] 9 is an explanatory diagram illustrating the treatment guideline model 44. The treatment guideline model 44 is a model that receives the objective variables output from the aforementioned renal condition model 42 and patient information including the patient's age, sex, height, weight, biochemical test results, medical history, and information on drugs administered using the syringe pump 34 and the infusion pump 35, and outputs a treatment guideline.

[0091] As described above, the objective variables output from the renal condition model 42 are renal condition indicators, such as an estimated value of renal medullary oxygen partial pressure or an estimated value of renal artery blood flow, etc. The treatment guideline model 44 may receive input of a dimensionless indicator obtained by non-dimensionalizing the objective variables output from the renal condition model 42 with an initial value.

[0092] The renal condition model 42 may receive input of a dependent variable calculated based on the multiple regression equation described in the first embodiment, instead of the dependent variable output from the renal condition model 42 .

[0093] The treatment guideline model 44 is generated by machine learning using an algorithm such as random forest or XGBoost. Training data containing numerous sets of objective variables, patient information, renal status indicators, treatment status, and treatment guidelines determined by specialists is used for the machine learning of the treatment guideline model 44. The treatment guideline is expressed by the type and dosage of a drug to be administered using the syringe pump 34 or the infusion pump 35, for example.

[0094] For example, by using training data in which a treatment guideline, etc., aimed at optimizing the balance between urine flow rate and urinary oxygen is recorded, a treatment guideline model 44 is generated that outputs a treatment guideline, etc., aimed at optimizing the balance between urine flow rate and urinary oxygen. For example, by using training data in which a treatment guideline, etc., aimed at making a renal status index a predetermined value is recorded, a treatment guideline model 44 is generated that outputs a treatment guideline, etc., aimed at making a renal status index a predetermined value.

[0095] The treatment guideline model 44 may be an algorithm such as a decision tree that is created on a rule-based basis in accordance with guidelines regarding standard treatment methods established by medical societies or individual medical institutions.

[0096] The treatment guideline model 44 may be re-learned using a database (DB) that records the treatment results of multiple patients. The re-learning may be performed by a large-scale computer or the like different from the information processing device 20, and the updated treatment guideline model 44 may be distributed to the information processing device 20 via a network. This makes it possible to realize an information processing system 10 that can constantly improve the accuracy of the treatment guideline model 44.

[0097] 10 is a flowchart illustrating the processing flow of the program according to embodiment 3. The processing up to step S507 is the same as the processing flow of the program according to embodiment 1 described using FIG. 2, and therefore the description thereof will be omitted.

[0098] The control unit 21 inputs the patient information obtained from the electronic medical record system 17, the syringe pump 34, and the infusion pump 35, and the renal condition index calculated in step S506 or step S507 into the treatment guideline model 44 to obtain a treatment guideline (step S521).

[0099] The control unit 21 displays the renal status index and the treatment guideline on the display unit 25 (step S522). The control unit 21 determines whether or not to end the process (step S509). The subsequent process is the same as the process flow of the program of the embodiment described using Figure 2, so a description thereof will be omitted.

[0100] 11 and 12 are example screens according to embodiment 3. The screen shown in Fig. 11 includes a renal status index field 72, a renal status indicator 721, a connection status field 71, and a patient information field 74, as well as two syringe pump fields 771, two infusion pump fields 772, and a suggestion field 78. In Fig. 11, no treatment guidelines are proposed, and the suggestion field 78 is blank.

[0101] The syringe pump field 771 displays the type of drug prepared in the syringe pump 34 and its setting status. "Syringe pump 1" is in a state where preparation for drug administration is complete. "Syringe pump 2" is in a state where preparation for drug administration is not complete. The infusion pump field 772 displays the type of drug prepared in the infusion pump 35 and the administration rate. In FIG. 11, right-pointing triangles indicate that infusion is currently being injected into the patient's body from "Infusion pump 1" and "Infusion pump 2."

[0102] 11, no treatment guideline is proposed, and the proposal column 78 is blank. In FIG. 12, a treatment guideline consisting of two items, "Stop infusion pump 2" and "Start syringe pump 1," is displayed in the proposal column 78. According to this embodiment, an information processing system 10 that displays a treatment guideline can be provided.

[0103] For example, when a doctor triple-tap the suggestion field 78, the control unit 21 may send control signals to the syringe pump 34 and the infusion pump 35 to execute the treatment guideline displayed in the suggestion field 78. It is possible to provide the information processing system 10 that allows a doctor to execute the displayed treatment guideline with simple instructions.

[0104] [Fourth Embodiment] This embodiment relates to an information processing system 10 that predicts changes in renal status indices. Explanation of parts common to the second embodiment will be omitted.

[0105] 13 is an explanatory diagram illustrating the renal condition model 42 according to the fourth embodiment. The renal condition model 42 according to the fourth embodiment receives explanatory variables and outputs a time-series dependent variable. That is, the renal condition model 42 according to the fourth embodiment outputs, in addition to the dependent variable at the current time, a predicted value for the dependent variable when the currently administered medication or other treatment is continued.

[0106] The renal condition model 42 is generated by machine learning using a training database that records many pairs of explanatory variables and time-series data of target variables. The explanatory variables may include time-series data such as time-series changes in vital data and time-series changes in medication dose. The renal condition model 42 of this embodiment uses an algorithm suitable for processing time-series data, such as LSTM (Long Short Term Memory) or Transformer. A reinforcement learning technique may also be used.

[0107] 14 and 15 are example screens according to the fourth embodiment. In the screen shown in Fig. 14, a predicted value field 79 is located to the right of the renal status index field 72. In Fig. 4, the predicted value field 79 displays the renal status index predicted value six hours later. When the triangle located to the right of "6h" is selected, the control unit 21 displays the predicted value of the renal status index 24 hours later.

[0108] 15, a renal status graph 75 is arranged below a renal status index column 72 and a predicted value column 79. The horizontal axis of the renal status graph 75 represents time, and the vertical axis of the renal status graph 75 represents a dimensionless renal status index.

[0109] The renal status graph 75 includes an actual value graph 752 and a predicted value graph 753. The actual value graph 752 displays actual data of the renal status index. The predicted value graph 753 plots predicted values ​​of the renal status index 6 hours, 24 hours, 48 ​​hours, and 72 hours later. Compared to the actual value graph 752, the predicted value graph 753 has a longer time per unit length on the horizontal axis.

[0110] 15 shows that if the current treatment is continued, the renal status index will decrease to nearly half of its current value after six hours. The suggestion field 78 displays, "We recommend reviewing treatment." The user can prevent acute kidney injury by avoiding significant changes in the renal status index, such as by changing medication.

[0111] In this embodiment, in steps S503 and S506 of the flowchart described using FIG. 2, the control unit 21 inputs explanatory variables into the renal state model 42 described using FIG. 8 to obtain time-series renal medullary oxygen partial pressures.

[0112] According to this embodiment, it is possible to provide an information processing system 10 that displays predicted values ​​of renal status indices.

[0113] Fifth Embodiment This embodiment relates to an information processing system 10 including a drug model 43 that estimates a future objective variable. Explanation of parts common to the first embodiment will be omitted.

[0114] FIG. 16 is an explanatory diagram illustrating the drug model 43. The drug model 43 is a trained model that receives the explanatory variables described using FIG. 8 and other figures and information related to a drug to be administered, and outputs a predicted value of an index at a predetermined time after the scheduled drug is administered. The index is a value that can be measured in real time, such as urinary oxygen partial pressure, urine flow rate, mean blood pressure, pulse pressure, heart rate, or blood oxygen saturation. The index may also be an index that can be estimated in real time, such as an estimated value of renal artery blood flow or an estimated value of renal tissue oxygen partial pressure. The drug model 43 may output multiple indexes.

[0115] The drug model 43 is generated by machine learning using training data acquired from, for example, animal experiments using pigs as described in embodiment 1. If the safety of the drug to be administered is sufficiently guaranteed, training data recording many sets of explanatory variables related to actual patients and post-dose indicators may be used. The drug model 43 may be a rule-based algorithm created based on the pharmacological action of the drug.

[0116] Responses to drugs vary from person to person, and even if a drug model 43 is used that is based on experimental data from pigs or data from a large number of past patients, it is difficult to accurately predict how each patient will respond to an administered drug.

[0117] During long surgeries and intensive care in the ICU, drugs that affect the kidneys may be administered repeatedly. By adjusting the drug dosage for each individual patient based on changes in the patient's physical condition after drug administration, sudden changes in the patient's condition can be avoided.

[0118] For example, if the change in the index after a predetermined time has elapsed after administration of a drug in accordance with the treatment guideline output from the treatment guideline model 44 described using Fig. 9 is greater than the prediction output from the drug model 43, it can be determined that the patient is the type of patient who responds strongly to the drug. The doctor can prevent kidney damage by reducing the drug dose from the next time onwards to a dose less than the amount output from the treatment guideline model 44.

[0119] The control unit 21 may add a coefficient calculated based on the difference between the predicted value of the index after drug administration output from the drug model 43 and the index actually measured a predetermined time after drug administration to the predicted value of the index output from the drug model 43 next time onwards, and display the result. The control unit 21 may re-learn the drug model 43 based on the difference between the predicted value of the index after drug administration output from the drug model 43 and the index actually measured after drug administration. As described above, an information processing system 10 can be provided that outputs indexes according to individual patients.

[0120] 17 is a flowchart illustrating the processing flow of the program of embodiment 5. The control unit 21 acquires urine information from the urine measurement device 31 (step S541). The control unit 21 acquires circulatory dynamics information from the vital sign monitor 33 (step S542). The control unit 21 acquires patient information from the electronic medical record system 17 (step S543). As described above, the patient information includes the patient's age, sex, height, weight, biochemical test results, medical history, and progress of ongoing treatment.

[0121] The control unit 21 acquires information about the drug to be administered (step S544). The information about the drug to be administered is acquired, for example, from the treatment guideline model 44 described with reference to Fig. 9. The control unit 21 may accept input of information about the drug to be administered by the user.

[0122] The control unit 21 inputs the explanatory variables and information about the drug scheduled to be administered to the patient into the drug model 43, and acquires a prediction index for a predetermined time after the drug is administered (step S545). The control unit 21 determines whether the drug was administered as scheduled in step S544 (step S546). The drug administration status is acquired, for example, from the syringe pump 34 or the infusion pump 35. The control unit 21 may also accept input from the user regarding the status of drug administration.

[0123] If it is determined that the drug was not administered as scheduled (NO in step S546), the control unit 21 returns to step S541. If it is determined that the drug was administered as scheduled (YES in step S546), the control unit 21 waits for a predetermined time to elapse in order to determine the effect of the drug (step S547).

[0124] The control unit 21 acquires the actual measurement value of the index based on the latest urine information or circulatory dynamics information (step S548). The control unit 21 calculates the difference between the predicted value acquired in step S545 and the actual measurement value acquired in step S548 (step S549). The control unit 21 determines whether the difference is equal to or smaller than a predetermined threshold (step S550).

[0125] If it is determined that the difference is not equal to or less than the threshold (NO in step S550), the control unit 21 corrects the drug model 43 so that the output data output from the drug model 43 approaches the actual measured value (step S551). Specifically, for example, the control unit 21 sets the display program to output a correction value obtained by multiplying the output of the drug model 43 by a coefficient obtained by dividing the actual measured value obtained in step S548 by the predicted value obtained in step S545. The control unit 21 may accept user input regarding a method for correcting the output of the drug model 43. The control unit 21 may re-learn the drug model 43.

[0126] If it is determined that the difference is equal to or less than the threshold value (YES in step S550), or after step S551 is completed, the control unit 21 determines whether or not to terminate the process (step S552). If it is determined not to terminate the process (NO in step S552), the control unit 21 returns to step S541. If it is determined to terminate the process (YES in step S552), the control unit 21 terminates the process.

[0127] According to this embodiment, it is possible to provide an information processing system 10 that predicts the efficacy of a drug for an individual patient.

[0128] 18 is an explanatory diagram illustrating the configuration of an information processing system 10 according to a sixth embodiment. This embodiment relates to a form in which an information processing device 20 is realized by combining and operating a general-purpose computer 90 and a program 97. Explanations of parts common to the first embodiment will be omitted.

[0129] The computer 90 includes a reading unit 29 in addition to the control unit 21, main memory device 22, auxiliary memory device 23, communication unit 24, display unit 25, input unit 26 and bus.

[0130] The program 97 is recorded on a portable recording medium 96. The control unit 21 reads the program 97 via the reading unit 29 and stores it in the auxiliary storage device 23. The control unit 21 may also read the program 97 stored in a semiconductor memory 98, such as a flash memory, implemented in the computer 90. Furthermore, the control unit 21 may download the program 97 from another server computer (not shown) connected via the communication unit 24 and a network (not shown) and store it in the auxiliary storage device 23.

[0131] The program 97 is installed as a control program for the computer 90, and is loaded into the main storage device 22 and executed. This realizes the information processing device 20 described in embodiment 1. The program 97 in this embodiment is an example of a program product.

[0132] The technical features (constituent elements) described in each embodiment can be combined with each other, and by combining them, new technical features can be formed. The embodiments disclosed herein are illustrative in all respects and should be considered not to be limiting. The scope of the present invention is defined by the claims, not by the meaning described above, and is intended to include all modifications within the meaning and scope of the claims.

[0133] 10 Information processing system 17 Electronic medical record system 20 Information processing device 21 Control unit 22 Main memory device 23 Auxiliary memory device 24 Communication unit 25 Display unit 26 Input unit 29 Reading unit 31 Urine measurement device 311 Urine sensor 33 Vital sign monitor 34 Syringe pump 35 Infusion pump 42 Renal status model 43 Drug model 44 Treatment guideline model 71 Connection status field 72 Renal status index field 721 Renal status indicator 731 Urinary oxygen partial pressure field 732 Urine flow field 733 Mean blood pressure field 734 Heart rate field 735 Arterial oxygen saturation field 736 Inhaled oxygen concentration field 74 Patient information field 75 Renal status graph 751 Range indicator 752 Actual value graph 753 Predicted value graph 761 Condition selection field 762 Correlation graph column 771 Syringe pump column 772 Infusion pump column 78 Proposal column 79 Prediction value column 90 Computer 96 Portable recording medium 97 Program 98 Semiconductor memory

Claims

1. Obtain urine information including urinary oxygen pressure, Calculating a renal status index based on the urine information; Outputting the renal status index A program that causes a computer to carry out processing.

2. The urine information includes at least one of the following in addition to the urinary oxygen pressure: urine volume, urine flow rate, urine color, urine absorbance, urine sodium content, and urine creatinine content. The program according to claim 1.

3. The renal condition index is an index related to an estimated value of renal artery blood flow, The urine information includes at least one of the urinary oxygen partial pressure and the urine flow rate, The renal condition index is calculated based on the urinary oxygen partial pressure and the urinary flow rate using a predetermined algorithm. The program according to claim 1.

4. The renal status index is an index related to an estimated renal tissue oxygen partial pressure; The urine information includes at least one of the urinary oxygen partial pressure and the urine flow rate, The renal condition index is calculated based on the urinary oxygen partial pressure and the urinary flow rate using a predetermined algorithm. The program according to claim 1.

5. The renal status index is an index related to an estimated value of glomerular filtration rate, The urine information includes at least one of the urinary oxygen partial pressure and the urine flow rate, The renal condition index is calculated based on the urinary oxygen partial pressure and the urinary flow rate using a predetermined algorithm. The program according to claim 1.

6. Outputting the urine information and vital data together with the renal condition index. The program according to claim 1.

7. Acquire circulatory dynamics information, The renal condition index is calculated based on the urine information and the circulatory dynamics information. The program according to claim 1.

8. The circulatory dynamics information includes at least one of mean blood pressure, pulse pressure, heart rate, blood oxygen saturation, and tissue oxygen saturation. The program according to claim 7.

9. The renal condition index is an index related to an estimated value of renal artery blood flow, The urine information includes at least one of the urinary oxygen partial pressure and the urine flow rate, The circulatory dynamics information includes a mean blood pressure, a pulse pressure, and a heart rate. Calculating the renal condition index using a predetermined algorithm based on the urinary oxygen partial pressure, the urine flow rate, the mean blood pressure, the pulse pressure, and the heart rate. The program according to claim 7.

10. The renal status index is an index related to an estimated renal tissue oxygen partial pressure; The urine information includes at least one of the urinary oxygen partial pressure and the urine flow rate, The circulatory dynamics information includes a mean blood pressure, a pulse pressure, and a heart rate. Calculating the renal condition index using a predetermined algorithm based on the urinary oxygen partial pressure, the urine flow rate, the mean blood pressure, the pulse pressure, and the heart rate. The program according to claim 7.

11. The renal status index is an index related to an estimated value of glomerular filtration rate, The urine information includes at least one of the urinary oxygen partial pressure and the urine flow rate, The circulatory dynamics information includes a mean blood pressure, a pulse pressure, and a heart rate. Calculating the renal condition index using a predetermined algorithm based on the urinary oxygen partial pressure, the urine flow rate, the mean blood pressure, the pulse pressure, and the heart rate. The program according to claim 7.

12. The urine information, the circulatory dynamics information, and the vital data are output together with the renal condition index. The program according to claim 7.

13. Repeatedly calculating the renal status index over time; The newly calculated renal status index is output as a relative value to the previously calculated renal status index. The program according to claim 1.

14. Obtain information regarding planned drug administration; Calculating a predicted value of the renal status index when the drug administration is performed; Output the predicted value The program according to claim 1.

15. calculating a predicted value of the renal status index after a predetermined time has elapsed; Output the predicted value The program according to claim 1.

16. Calculating a difference between the predicted value and the renal status index at the time when the predetermined time has elapsed; The difference is used to calculate a predicted value of the future renal status index. The program according to claim 15.

17. Outputting a treatment guideline based on the renal condition index 17. A program according to any one of claims 1 to 16.

18. The treatment guideline is a treatment guideline aimed at bringing the relationship between urine flow rate and urinary oxygen partial pressure into a predetermined state.

18. The program according to claim 17.

19. The treatment guideline is a treatment guideline aimed at bringing the renal status index to a predetermined value.

18. The program according to claim 17.

20. Injecting medicine into the medicine injection device based on the treatment guideline.

18. The program according to claim 17.

21. Repeatedly calculating the renal status index over time; changing the amount of drug administered by the drug administration device based on a change in the renal status index after the drug administration device has administered the drug; The program according to claim 20.

22. When a renal status index and a treatment guideline are input, a trained model is used to output a renal status index after a drug administration based on the treatment guideline is performed, to obtain a predicted value of the renal status index after a drug administration based on the treatment guideline is performed; Calculating a difference between the predicted value and an actual renal status index after drug administration based on the treatment guideline is performed; Re-learning the trained model based on the difference.

18. The program according to claim 17.

23. When a renal status index and a treatment guideline are input, a trained model is used to output a renal status index after a drug administration based on the treatment guideline is performed, to obtain a predicted value of the renal status index after a drug administration based on the treatment guideline is performed; Calculating a difference between the predicted value and an actual renal status index after drug administration based on the treatment guideline is performed; Adjusting the output data of the trained model based on the difference 18. The program according to claim 17.

24. Obtain urine information including urinary oxygen pressure, Calculating a renal status index based on the urine information; Outputting the renal status index An information processing method in which processing is performed by a computer.

25. An information processing device including a control unit, The control unit Obtain urine information including urinary oxygen pressure, Calculating a renal status index based on the urine information; Outputting the renal status index Information processing device.