JP2021108908A - Dialysis support device, dialysis support method, and dialysis support program - Google Patents

Abstract

To reduce costs of dialysis treatment while maintaining safety of the dialysis treatment.SOLUTION: A dialysis support device 10 includes: an acquisition unit for acquiring dialysis condition data on a specific patient; a prediction unit for generating a prediction value of blood pressure of the specific patient by inputting the acquired dialysis condition data to a blood pressure prediction model generated on the basis of machine learning using dialysis condition data on each patient and a blood pressure value of each patient measured when executing dialysis treatment using the dialysis condition data; and a notification unit for notifying of warning information on the specific patient on the basis of the prediction value.SELECTED DRAWING: Figure 1

Description

The present invention relates to a dialysis support device, a dialysis support method, and a dialysis support program.

Conventionally, dialysis treatment such as hemodialysis (HD) is known as a treatment method for kidney diseases such as renal failure. In dialysis treatment, blood taken out of the body is passed through a dialyzer (also called a dialyzer) to remove waste products in the blood, remove excess water (also called water removal, etc.), and adjust electrolytes and acidity. By doing at least one of the above, the washed blood is returned to the body again.

In dialysis treatment, in general, the conditions for dialysis treatment of a patient (for example, scheduled water removal time, prescribed drug, water removal amount, water removal rate, etc., hereinafter also referred to as dialysis conditions) are determined by a doctor's examination, and a medical record is obtained. Created. Medical professionals such as nurses and technicians set dialysis conditions in the dialysis machine according to the contents of the medical record from the doctor, and perform dialysis treatment for the patient. As a device for supporting such dialysis treatment, for example, Patent Document 1 describes a dialysis support device that automatically creates a medical record used for dialysis treatment.

Japanese Unexamined Patent Publication No. 2018-175282

In dialysis treatment, a patient's blood pressure is likely to decrease due to the influence of water removal and the like. Therefore, it is important in dialysis treatment to set dialysis conditions so that the patient's blood pressure can be appropriately maintained and to appropriately monitor changes in the patient's blood pressure during hemodialysis.

However, what kind of dialysis condition is determined often depends on the experience of the doctor, and even if the dialysis condition specified by the doctor is used, it is not always possible to appropriately maintain the blood pressure of the patient during the dialysis treatment. In addition, if the patient's blood pressure drops sharply during dialysis treatment and the patient falls into a dangerous state, urgent measures such as changing the setting of the dialysis device are required. For this reason, a healthcare professional who monitors the blood pressure of patients undergoing dialysis treatment is required, but increasing the ratio of the healthcare professional to the number of beds improves the safety of patients undergoing dialysis treatment, while improving the safety of the patients. May increase the cost of dialysis treatment.

Therefore, one of the objects of the present inventor is to provide a dialysis support device, a dialysis support method, and a dialysis support program capable of reducing the cost of dialysis treatment while maintaining the safety of dialysis treatment.

The dialysis support device according to one aspect of the present invention was measured when performing a dialysis treatment using the dialysis condition data of each patient and the dialysis condition data, and the acquisition unit that acquires the dialysis condition data of a specific patient. With the prediction unit that generates the predicted value of the blood pressure of the specific patient by inputting the acquired dialysis condition data into the blood pressure prediction model generated based on the machine learning using the blood pressure value of each patient. , A notification unit for notifying warning information about the specific patient based on the predicted value.

According to this aspect, since the warning information is notified based on the predicted value, it becomes easy for the medical staff to recognize the sudden decrease in blood pressure of the patient in advance. Therefore, the ratio of the medical staff to the number of beds can be reduced while maintaining the safety of the dialysis treatment, and as a result, the cost of the dialysis treatment can be reduced.

In the above aspect, in the machine learning, in addition to the dialysis condition data and the blood pressure value, the medical care data of each patient is used, and the acquisition unit acquires the medical care data of the specific patient and predicts the above. The value may be generated by inputting the dialysis condition data and the medical care data into the blood pressure prediction model.

According to this aspect, a blood pressure prediction model generated based on machine learning using the medical data of each patient in addition to the dialysis condition data and the blood pressure value is used, so that the accuracy of the predicted value is improved. be able to.

In the above aspect, the notification unit may notify the warning information when the predicted value is equal to or less than a predetermined threshold value or smaller than the predetermined threshold value. According to this aspect, the warning information can be notified more appropriately.

In the above aspect, the predicted value generated at a second time interval shorter than the first time interval at which blood pressure is actually measured when performing dialysis treatment for the specific patient may be notified. According to this aspect, when performing dialysis treatment, a predicted value is generated at a second time interval shorter than the first time interval at which the blood pressure is actually measured, so that the possibility of a sudden decrease in blood pressure of the patient can be quickly increased. Can be recognized. As a result, it can contribute to the safety of patients undergoing dialysis treatment.

The dialysis support method according to another aspect of the present invention was measured in the step of acquiring dialysis condition data of a specific patient and in performing dialysis treatment using the dialysis condition data of each patient and the dialysis condition data. A step of inputting the acquired dialysis condition data into a blood pressure prediction model generated based on machine learning using the blood pressure values of each patient to generate a predicted value of the blood pressure of the specific patient. It has a step of notifying warning information about the specific patient based on the predicted value.

The dialysis support program according to another aspect of the present invention uses a calculation unit provided in the dialysis support device, an acquisition unit for acquiring dialysis condition data of a specific patient, dialysis condition data of each patient, and the dialysis condition data. By inputting the acquired dialysis condition data into a blood pressure prediction model generated based on machine learning using the blood pressure values of each patient measured when performing the dialysis treatment, the specific patient It functions as a prediction unit that generates a prediction value of blood pressure, and a notification unit that notifies warning information about the specific patient based on the prediction value.

According to the present invention, the cost of dialysis treatment can be reduced while maintaining the safety of dialysis treatment.

It is a figure which shows an example of the schematic structure of the dialysis support system which concerns on this embodiment. It is a figure which shows an example of the functional structure of the dialysis support system which concerns on this embodiment. It is a figure which shows an example of the hardware configuration of each apparatus constituting the dialysis support system which concerns on this embodiment. It is a flowchart which shows an example of the learning operation which concerns on this embodiment. It is a flowchart which shows an example of the dialysis support operation which concerns on this embodiment. It is a figure which shows an example of the relationship between the dialysis time and the water removal rate which concerns on this embodiment. It is a figure which shows an example of the relationship between the dialysis time and the water removal rate which concerns on this embodiment. It is a figure which shows an example of the display screen of the predicted value of blood pressure and warning information which concerns on this embodiment.

Embodiments of the present invention will be described with reference to the accompanying drawings. In each figure, those having the same reference numerals have the same or similar configurations.

(Configuration of dialysis support system)
<Outline configuration>
FIG. 1 is a diagram showing an example of a schematic configuration of a dialysis support system according to an embodiment of the present invention. As shown in FIG. 1, in the dialysis support system 1, a dialysis device 50 used for dialysis treatment for a patient and a dialysis support device 10 for supporting dialysis treatment using the dialysis device 50 are connected via a network. There is.

First, dialysis treatment using the dialysis apparatus 50 will be described. Dialysis treatment may include at least one of hemodialysis (HD) and extracorporeal ultrafiltration (ECM). HD is the indirect contact between blood and dialysate via a dialyzer (also referred to as a dialyzer or the like) to remove waste products accumulated in the body by the principle of diffusion. ECUM is to remove excess water in the body (also referred to as water removal or the like).

As shown in FIG. 1, the dialysis apparatus 50 controls the inflow / outflow of blood and the inflow / outflow of dialysate into and out of the dialyzer 51 (also referred to as a dialyzer or the like). The dialyzer 51 includes, for example, a semipermeable membrane made of a hollow fiber membrane, removes waste products (toxic substances) and excess water contained in blood, and replenishes deficient nutrients.

The dialyzer 51 includes a blood circuit through which patient blood circulates (not shown), a dialysate circuit through which dialysate circulates (not shown), and a replacement fluid circuit used for dialysis (not shown). A blood pump 53 for circulating blood from a patient is connected to the blood inlet 52 of the dialyzer 51 via an arterial drip chamber 54. A venous drip chamber 56 is provided at the blood outlet 55 of the dialyzer 51.

In the blood circuit, the blood pump 53 introduces the patient's blood from the artery into the dialyzer 51 via the arterial drip chamber 54 and then back into the vein via the venous drip chamber 56. In the dialysate circuit, the dialysate inlet 57 of the dialyzer 51 is connected to the dialyzer 50 via the dialysate pump 58, and fresh dialysate is supplied to the dialyzer 51. The dialysing device 50 is also connected to the dialysate outlet 59 of the dialyzer 51, and the used dialysate is discharged to the dialysing device 50. In the fluid replacement circuit, fluid replacement (for example, a coagulant) is supplied from the dialysis apparatus 50 to the arterial drip chamber 54, and is supplied to the patient's blood in the dialyzer 51.

Further, in the dialyzer 51, excess water in the blood may be removed (water removal) by the ECUM. When the amount of water in the blood vessel decreases due to the water removal, the water outside the blood vessel moves into the blood vessel, and a reaction (also referred to as plasma refilling) that tries to maintain the amount of water in the blood vessel occurs. In plasma refilling, water in the patient's body may move from the intracellular to the extracellular (stroma) and from the stroma into the blood vessels. Such movement of water is caused by a decrease in intravascular pressure (hydrostatic pressure) due to water removal, a force for drawing water by proteins in blood (oncotic pressure), and the like.

Plasma refilling can avoid excessive reductions in blood volume and blood pressure during dialysis treatment. For patients who are less likely to undergo plasma refilling (eg, hypoproteinemia), certain drugs (eg, hypertonic sodium chloride solution, concentrated glycerin / fructose injection, etc.) that increase the osmotic pressure in the blood. ) May be used to promote plasma refilling and stabilize blood pressure.

The dialysis apparatus 50 is set with patient dialysis conditions (for example, type of dialysis machine 51, drug, dialysis time, blood flow velocity, dry weight, etc.). The dialysis apparatus 50 controls the amount of dialysate sent to the dialyzer 51, the amount of dialysate (replenishment) sent to the arterial drip chamber 54 through the replacement fluid circuit, and the like according to the set dialysis conditions. The dry weight is a body weight in which excess water is not accumulated in the body (that is, after dialysis), and is also called an appropriate body weight, a reference body weight, or the like.

The dialysis support device 10 predicts the blood pressure of a patient who undergoes dialysis treatment using the dialysis device 50 based on the data regarding the dialysis conditions (dialysis condition data) set in the dialysis device 50. The dialysis support device 10 may be connected to the learning device 20 that performs learning processing of the blood pressure prediction model used for predicting the blood pressure via a network.

Further, the dialysis support system 1 is a database (DB) that stores data related to dialysis treatment using the dialysis apparatus 50 (for example, the above dialysis condition data, blood pressure values measured when performing the dialysis treatment, etc.). Dialysis DB 30 may be included. As shown in FIG. 1, the dialysis DB 30 may be connected to the dialysis device 50 and the learning device 20 via a network.

Further, the dialysis support system 1 may include a medical care DB 40 which is a database (DB) for storing data (medical care data) related to medical care of a patient who performs dialysis treatment using the dialysis apparatus 50. As shown in FIG. 1, the medical care DB 40 may be connected to the dialysis support device 10 and the learning device 20 via a network.

According to the dialysis support system 1 as described above, the blood pressure of the patient who performs the dialysis treatment can be predicted by using the dialysis device 50, so that the medical staff can determine the validity of the dialysis conditions set in the dialysis device 50 in advance. You can check. The prediction also allows healthcare professionals to be aware in advance of sudden changes in the patient's blood pressure during dialysis treatment and to take action before the patient is in danger.

The schematic configuration shown in FIG. 1 is merely an example, and is not limited to the illustrated one. For example, although FIG. 1 shows a case where the dialysis support device 10 and the learning device 20 are separate bodies, the dialysis support device 10 and the learning device 20 may be configured as an integrated device. For example, the dialysis support device 10 and the learning device 20 may be realized as different operations of the PLC (Programmable Logic Controller). Further, the dialysis DB 30 and the medical care DB 40 may be replaced with a computer-readable non-temporary storage medium (for example, a DVD). Further, when the blood pressure prediction model by distillation described later is used, the learning device 20 may not be included. Further, one or more dialysis devices 50 (including each configuration designated by reference numerals 51 to 59) may be connected to the dialysis support device 10. Further, the dialysis DB 30 may be connected to the dialysis support device 10.

<Functional configuration>
≪Learning device≫
FIG. 2 is a diagram showing an example of the functional configuration of the dialysis support system according to the embodiment of the present invention. As shown in FIG. 2, the learning device 20 has a dialysis condition data acquisition unit 21, a medical care data acquisition unit 22, a blood pressure value acquisition unit 23, a storage unit 24, and a learning unit 25 as functional configurations. ..

The dialysis condition data acquisition unit 21 acquires dialysis condition data for each patient. Specifically, the dialysis condition data acquisition unit 21 may acquire the dialysis condition data of each patient stored in the dialysis DB 30. Here, the dialysis condition data is data relating to the hemodialysis conditions of each patient, and may include at least one of the following for each patient, for example.
・ Information on blood volume (BV) (BV information)
-Information (PRR information) regarding the above-mentioned plasma refilling rate (plasma refilling rate (PRR)).
・ Information about blood tests (blood test information)
・ Patient information (patient information)

Here, the BV information may indicate the circulating blood volume in the patient's body. For example, the patient's systemic circulating blood volume may be indicated by 8% (1/13) of the patient's body weight.

Further, the PRR information may be information on the body fluid volume of each patient (body fluid volume information) and / or information on the osmotic pressure (osmotic pressure information), and includes, for example, information on at least one of the following for each patient. It may be.
-Water removal rate-Dialysis time (The dialysis time may be, for example, the time related to the ECUM and / or HD.)
-Dialysis amount (Note that the dialysis amount may be indicated by kt / V. Here, k is the amount of body fluid (also referred to as clearance) without toxin (urea) produced within a predetermined time, and t is It is the dialysis time, and V may be the total fluid volume (eg, 60% of each patient's body weight).
-Weight of each patient (eg, weight before and / or after hemodialysis, dry weight, etc.)
・ Cumulative amount of water removed for each dialysis time ・ Cardiothoracic ratio ・ Echocardiography (left ventricular end-diastolic dimension (LVDd))
・ Serum albumin / hemoglobin (hemoglobin (Hb))

The blood test information is a blood test value before and / or after hemodialysis, and may include, for example, a value relating to at least one of the following.
・ Blood urea nitrogen (BUN)
・ Creatinine
・ Low-density lipoprotein cholesterol (LDL-C)
・ Triglyceride
・ Total protein
・ C-reactive protein
・ White blood cell count
・ Hematocrit (Hct)
・ Hemoglobin (hemoglobin (Hb))
・ Platelets (plaque (Plt))

Further, the patient information is information about the patient, and may include, for example, information about at least one of the following.
・ Name ・ Date of birth ・ Age ・ Gender ・ Presence or absence of diabetes (Diabetes) ・ Value of nutrition-related index (for example, geriatric nutritional risk index (GNRI)) ・ Presence or absence of treatment of dyslipidemia ・Heart-rate

The medical care data acquisition unit 22 acquires medical care data of each patient. Specifically, the medical care data acquisition unit 22 may acquire medical care data of each patient stored in the medical care DB 40. Here, the medical care data is data related to the medical care of each patient, and may include at least one of the following for each patient, for example. The medical data may be called an electronic medical record or the like.
・ Information on medical history ・ Information on current medical history (for example, presence or absence of diabetes, presence or absence of treatment for dylipidemia, etc.)
・ Information about medicine ・ Name ・ Date of birth ・ Age ・ Gender

The blood pressure value acquisition unit 23 acquires the blood pressure value of each patient measured when performing dialysis treatment using the dialysis condition data in the dialysis condition data acquisition unit 21. Specifically, the blood pressure value acquisition unit 23 may acquire the blood pressure value of each patient stored in the dialysis DB 30. The blood pressure value may be measured before the dialysis treatment of each patient, at a predetermined time point during the dialysis treatment (for example, at a predetermined time point), after the dialysis treatment, or the like.

The storage unit 24 stores the learning data 24a and the blood pressure prediction model 24b generated by the learning process by the learning unit 25. The learning data 24a may be data in which the dialysis condition data acquired by the dialysis condition data acquisition unit 21 and the blood pressure value acquired by the blood pressure value acquisition unit 23 are associated with each patient. Further, the learning data 24a may be data in which, in addition to the dialysis condition data and the blood pressure value, the medical care data acquired by the medical care data acquisition unit 22 is associated with each patient.

The learning unit 25 executes the learning process of the blood pressure prediction model 24b by machine learning using the learning data 24a stored in the storage unit 24. For example, the learning unit 25 of the blood pressure prediction model 24b so that the blood pressure value of each patient stored as the learning data 24a is output based on the dialysis condition data of each patient stored as the learning data 24a. Learning processing may be performed. Further, the learning unit 25 predicts the blood pressure so that the blood pressure value of each patient stored as the learning data 24a is output based on the dialysis condition data and the medical care data of each patient stored as the learning data 24a. The learning process of the model 24b may be performed.

The blood pressure prediction model 24b machine-learned by the learning unit 25 may be composed of, for example, a neural network. When the blood pressure prediction model 24b is composed of a neural network, the learning unit 25 may perform the learning process of the neural network by, for example, the error back propagation method. also. The blood pressure prediction model 24b may be constructed based on deep learning, which is a neural network having two or more intermediate layers.

≪Dialysis support device≫
The dialysis support device 10 includes a dialysis condition data acquisition unit 11, a medical care data acquisition unit 12, a prediction unit 13, and a notification unit 14. The dialysis condition data acquisition unit 11 acquires dialysis condition data of a specific patient who performs dialysis treatment with the dialysis apparatus 50. The medical care data acquisition unit 12 acquires medical care data of the specific patient.

The details of the dialysis condition data acquisition unit 11 and the medical care data acquisition unit 12 are as described in the dialysis condition data acquisition unit 21 and the medical care data acquisition unit 22, respectively. As shown in FIG. 2, the number of dialysis machines 50 may be one or more, and one dialysis machine 50 may be set for each bed. Further, the dialysis condition data set in the dialysis apparatus 50 and the blood pressure value measured when performing the dialysis treatment using the dialysis condition data may be sequentially stored in the dialysis DB 30. The dialysis condition data acquisition unit 11 may acquire the dialysis condition data set in the dialysis apparatus 50, or may acquire the dialysis condition data of the specific patient stored in the dialysis DB 30.

The prediction unit 13 inputs the dialysis condition data acquired by the dialysis condition data acquisition unit 11 into the blood pressure prediction model 13a to generate a predicted value of the patient's blood pressure. Further, the prediction unit 13 may input the medical care data acquired by the medical care data acquisition unit 12 into the blood pressure prediction model 13a in addition to the dialysis condition data to predict the blood pressure value of the patient.

The blood pressure prediction model 13a may be a blood pressure prediction model 24b generated by machine learning in the learning device 20, or a new blood pressure prediction generated by "distillation" in machine learning or generated by distillation. It may be a model. In "distillation", a new prediction model may be generated using the input / output data to the blood pressure prediction model 24b machine-learned by the learning device 20. The blood pressure prediction model 13a may be paraphrased as a prediction program, a prediction algorithm, a predictor, or the like.

The notification unit 14 notifies warning information about a patient undergoing dialysis treatment using the dialysis apparatus 50 based on the predicted value of blood pressure generated by the prediction unit 13. Specifically, the notification unit 14 may output the warning information to the output unit 10e, which will be described later, display it on a predetermined display screen, and / or sound a warning sound. Alternatively, the notification unit 14 may transmit the warning information to another device (for example, each dialysis device 50) via the communication unit 10c. The other device may display the warning information on the display screen.

Further, the notification unit 14 may notify the predicted value of the blood pressure generated by the prediction unit 13. Specifically, the notification unit 14 may output the predicted value to the output unit 10e, which will be described later, and display it on a predetermined display screen. Alternatively, the notification unit 14 may transmit the predicted value to another device via the communication unit 10c. The predicted value may be generated at a second time interval, which is shorter than the first time interval at which blood pressure is actually measured when performing dialysis treatment for a specific patient.

<Hardware configuration>
Next, the hardware configuration of each device in the dialysis support system 1 (for example, at least one of the dialysis support device 10, the learning device 20, and the dialysis device 50) will be described. As shown in FIG. 3, each device in the dialysis support system 1 includes a CPU (Central Processing Unit) 10a corresponding to an arithmetic unit, a storage device 10b, a communication unit 10c, an input unit 10d, and an output unit 10e. Has. Each of these configurations is connected to each other via a bus so that data can be transmitted and received. In this example, the dialysis support device 10 and the learning device 20 are composed of separate computers, but may be composed of one computer.

The CPU 10a is a control unit that controls execution of a program stored in the storage device 10b, calculates data, and processes data. The CPU 10a may be an arithmetic unit (arithmetic unit) that executes a program (dialysis support program) that generates a predicted value of blood pressure of the patient based on the patient's dialysis condition data. The CPU 10a receives various input data from the input unit 10d and / or the communication unit 10c, outputs (for example, displays) the calculation result of the input data to the output unit 10e, stores it in the storage device 10b, or communicates. It may be transmitted via unit 10c.

The storage device 10b is at least one of a memory, an HDD (Hard Disk Drive), and an SSD (Solid State Drive). The storage device 10b of the learning device 20 may constitute a storage unit 24. The storage device 10b of the dialysis support device 10 may store the dialysis support program executed by the CPU 10a and the blood pressure prediction model 13a.

The communication unit 10c is an interface for connecting each device in the dialysis support system 1 to an external device. When the dialysis support device 10 and the learning device 20 are configured as an integrated device, the communication unit 10c performs interprocess communication between the process operating as the dialysis support device 10 and the process operating as the learning device 20. May include.

The input unit 10d receives data input from the user, and may include at least one of a keyboard, a mouse, a touch panel, and a microphone, for example.

The output unit 10e outputs the calculation result by the CPU 10a, and may be composed of at least one of a display such as an LCD (Liquid Crystal Display) and a speaker, for example.

The dialysis support program may be stored in a storage medium readable by a computer such as the storage device 10b and provided, or may be provided via a network connected by the communication unit 10c. The storage medium in which the dialysis support program is stored may be a non-transitory computer readable medium. The non-temporary storage medium is not particularly limited, but may be, for example, a storage medium such as a USB memory, a CD-ROM, or a DVD.

In the dialysis support device 10, when the CPU 10a executes the dialysis support program, the operations of the dialysis condition data acquisition unit 11, the medical care data acquisition unit 12, the prediction unit 13, and the notification unit 14 described with reference to FIG. 2 are realized. .. It should be noted that these physical configurations are examples and do not necessarily have to be independent configurations. For example, the dialysis support device 10 may include an LSI (Large-Scale Integration) in which the CPU 10a and the storage device 10b are integrated.

Further, the dialysis condition data acquisition units 11 and 21, the medical care data acquisition units 12 and 22, and the blood pressure value acquisition unit 23 may acquire the dialysis condition data, the medical care data, and the blood pressure value received by the communication unit 10c, respectively. Alternatively, the dialysis condition data, the medical treatment data, and the blood pressure value stored in a storage medium readable by a computer such as the storage device 10b may be acquired.

(Operation of dialysis support system)
<Learning movement>
FIG. 4 is a flowchart showing an example of the learning operation according to the present embodiment. The flowchart shown in FIG. 4 is not limited to this, and is not limited to this. For example, the medical data in steps S101 and S102 may be omitted.

As shown in FIG. 4, in step S101, the learning device 20 acquires dialysis condition data and medical care data of each patient and a blood pressure value of each patient. The dialysis condition data may be dialysis condition data actually used in the past dialysis treatment. Further, the blood pressure value may be a blood pressure value (also referred to as an actually measured value, a measured value, etc.) actually measured at the time of dialysis treatment using the dialysis condition data. For example, the learning device 20 may acquire the dialysis condition data and the blood pressure value stored in the dialysis DB 30. Further, the learning device 20 may acquire the medical care data stored in the medical care DB 40.

In step S102, the learning device 20 generates learning data 24a that associates the dialysis condition data and medical care data acquired in step S101 with the blood pressure value for each patient.

In step S103, whether or not the learning device 20 satisfies a predetermined condition (for example, whether or not the learning data 24a is sufficiently accumulated, or whether or not it is the timing to update the blood pressure prediction model 24b, etc. ) Is determined. When a predetermined condition is satisfied (step S103; Yes), this operation proceeds to step S104. On the other hand, if the predetermined condition is not satisfied (step S103; No), this operation returns to step S101.

In step 104, the learning device 20 may perform the learning process of the blood pressure prediction model 24b by machine learning using the learning data 24a generated in step S102. The determination in step S103 may be omitted.

<Dialysis support operation>
FIG. 5 is a diagram showing an example of a dialysis support operation according to the present embodiment. The flowchart shown in FIG. 5 is not limited to this, and is not limited to this. For example, the medical data in steps S201 and S202 may be omitted.

As shown in FIG. 5, in step S201, the dialysis support device 10 acquires dialysis condition data and medical care data of a specific patient. Specifically, the dialysis support device 10 may acquire dialysis condition data of a specific patient set in the dialysis device 50. Further, the dialysis support device 10 may acquire medical care data of the specific patient from the medical care DB 40.

In step S202, the dialysis support device 10 inputs the dialysis condition data and the medical care data acquired in step S201 into the blood pressure prediction model 13a, and uses the predicted value of the blood pressure of the specific patient as the output of the blood pressure prediction model 13a. Generate. The predicted value may be generated at a second time interval, which is shorter than the first time interval at which blood pressure is actually measured when performing dialysis treatment for the specific patient. The dialysis support device 10 may output the generated predicted value to the output unit 10e and display it on the display screen, or to another device (for example, each dialysis device 50) via the communication unit 10c. You may send it.

The predicted value generated in step S202 may be a predicted value of blood pressure after a predetermined time (for example, 10 minutes) from a predetermined time when the main dialysis support operation is performed. In this case, the dialysis condition data acquired in step S201 may be the dialysis condition data used after a predetermined time from the predetermined time point.

In step S203, the dialysis support device 10 determines whether or not the predicted value generated in step S202 satisfies a predetermined condition. The predetermined condition may be, for example, that the predicted value of blood pressure is equal to or less than a predetermined threshold value or smaller than the predetermined threshold value. The predetermined threshold value may be a value common to all patients or a value determined for each patient.

When the predicted value of the blood pressure satisfies the predetermined condition (step S203; Yes), the dialysis support device 10 may notify the warning information in step S204. Specifically, the dialysis support device 10 may output the warning information to the output unit 10e, display it on a predetermined display screen, and / or sound a warning sound. Alternatively, the dialysis support device 10 may transmit the warning information to another device (for example, each dialysis device 50) via the communication unit 10c. The other device may display the warning information on the display screen.

The above dialysis support operation may be performed before the dialysis treatment using the dialysis apparatus 50 is performed. When the main dialysis support operation is performed before the dialysis treatment is performed, the medical staff can confirm in advance whether the dialysis condition data set in the dialysis device 50 is appropriate for maintaining the blood pressure of the patient. Therefore, the safety of dialysis treatment can be maintained without much reliance on the experience of the doctor.

Alternatively, the above dialysis support operation may be performed at a predetermined time during the dialysis treatment using the dialysis device 50. The predetermined time point may be a time point (time point of a predetermined cycle) at each predetermined time interval (for example, 5 minutes, 10 minutes, etc.) within the dialysis time. The dialysis support operation during dialysis treatment will be described with reference to FIGS. 6A, 6B and 7.

6A and 6B are diagrams showing an example of the relationship between the dialysis time and the water removal rate according to the present embodiment. In FIGS. 6A and 6B, the dialysis time may be, for example, 4 hours (240 minutes). 6A and 6B show changes in the water removal rate during the dialysis time.

As in pattern 1 of FIG. 6A, the water removal speed may be gradually increased with the start time as the minimum speed and the end time as the maximum speed. Alternatively, as in pattern 2, the start time may be the maximum speed, the end time may be the minimum speed, and the water removal speed may be gradually slowed down. Alternatively, as in pattern 3, the water removal rate may be constant. Further, as in pattern 4 of FIG. 6B, the water removal rate may gradually increase from the start time point to the intermediate point point, reach the maximum speed at the intermediate point point, and gradually decrease from the intermediate point point to the end point point. Alternatively, as in pattern 5, the water removal rate may gradually decrease from the start point to the intermediate point, become the lowest rate at the intermediate point, and gradually increase from the intermediate point to the end point.

The pattern of the water removal rate as shown in FIGS. 6A and 6B is determined for each patient and may be set in the dialysis apparatus 50 as dialysis condition data. Further, the minimum speed and / or the maximum speed in each pattern may be set in the dialysis apparatus 50 in common among the patients, or may be set in the dialysis apparatus 50 for each patient.

In this way, when the water removal rate is controlled within the dialysis time when the dialysis treatment is performed, by performing the main dialysis support operation at predetermined time intervals, the patient suffers from a sudden change in blood pressure due to the change in the water removal rate. Can be prevented from falling into a dangerous state.

FIG. 7 is a diagram showing an example of a display screen of a predicted value of blood pressure and warning information according to the present embodiment. In FIG. 7, the output unit of the dialysis support device 10 when the dialysis support operation shown in FIG. 5 is performed at a predetermined time during dialysis treatment using each dialysis device 50 (here, seven dialysis devices 50). An example of the display screen D of 10e is shown. Note that FIG. 7 is merely an example, and the configuration of the display screen D is not limited to that shown. In addition, the predicted value and / or warning information may be displayed on another device (for example, each dialysis device).

As shown in FIG. 7, the display screen D may show a predicted value of blood pressure of each patient using each dialysis apparatus 50. For example, in FIG. 7, the dialysis apparatus 50 is arranged in each of the seven beds 001 to 007, and the dialysis treatment of patients A to G is performed using the dialysis apparatus 50 in each of the seven beds 001 to 007. And.

In the area R1 of the display screen D, predicted values of blood pressure of patients A to G after a predetermined time (for example, 10 minutes) from a specific time point (for example, 12:38 pm) within the dialysis time are displayed. For example, in region R1, predicted values of blood pressure of patients A to G after a predetermined time from a specific time point are displayed in ascending order. In FIG. 7, the area R1 is arranged at the right end of the display screen D, but the position of the area R1 is not limited to that shown in the drawing.

Further, the predicted values of patients A and F whose predicted values satisfy the above-mentioned predetermined conditions may be colored as warning information different from those of other patients B to E and G, for example. For example, in FIG. 7, the predetermined condition may be that the predicted value is equal to or less than or smaller than the threshold value T (here, 110 mmhg). The threshold T may be set individually for the patient A, or may be commonly set for the patients A to G.

Further, in the area R2 of the display screen D of FIG. 7, a change in the predicted value and / or the measured value of the blood pressure value within the dialysis time of a specific patient may be shown. For example, FIG. 7 shows changes in the predicted and measured blood pressure values of patient A during the dialysis time. In FIG. 7, black circles indicate actually measured values of blood pressure at a predetermined time point. As shown by the black circle in FIG. 7, the blood pressure may be measured at predetermined time intervals (for example, every 30 minutes, also referred to as a first time interval, etc.).

Further, in FIG. 7, the alternate long and short dash line indicates the change in the predicted value of blood pressure. For example, in FIG. 7, the predicted value of the time point T'is a predicted value of blood pressure at a predetermined time (for example, 10 minutes later) from the time point T predicted by the dialysis support device 10 at the time point T. In FIG. 7, when the predicted value of blood pressure of patient A is equal to or less than the threshold value T or smaller than the threshold value T, the dialysis support device 10 may notify the warning information. Further, as shown in FIG. 7, the probability of the predicted value (here, 92%) may be displayed in the region R1.

As shown in FIG. 7, even if the predicted value of the time point T is higher than the threshold value T, the predicted value of the time point T'after a predetermined time from the time point T is sharply lowered from the predicted value of the time point T. In this case, by recognizing the predicted value of the time point T'in advance, the medical staff can take measures to maintain the safety of the patient.

As described above, according to the dialysis support operation according to the present embodiment, the sudden decrease in blood pressure of the patient can be recognized by the warning information, so that the ratio of the medical staff to the number of beds can be determined while maintaining the safety of the dialysis treatment. Can be reduced. As a result, the cost of dialysis treatment can be reduced.

In the above embodiment, it is assumed that the dialysis apparatus 50 performs the HD and the ECUM together in one dialysis treatment, but the present invention is not limited to this. The dialysis apparatus 50 can be appropriately applied to blood pressure prediction in the case where the ECUM is performed without performing the HD in one dialysis or when the HD is performed without performing the ECUM.

The embodiments described above are for facilitating the understanding of the present invention, and are not for limiting and interpreting the present invention. Each element included in the embodiment and its arrangement, material, condition, shape, size, and the like are not limited to those exemplified, and can be changed as appropriate. In addition, the configurations shown in different embodiments can be partially replaced or combined.

1 ... Dialysis support system, 10 ... Dialysis support device, 20 ... Learning device, 30 ... Dialysis DB, 40 ... Medical DB, 50 ... Dialysis device, 11 ... Dialysis condition data acquisition unit, 12 ... Medical data acquisition unit, 13 ... Prediction Department, 14 ... Notification unit, 21 ... Dialysis condition data acquisition unit, 22 ... Medical data acquisition unit, 23 ... Blood pressure value acquisition unit, 24 ... Storage unit, 25 ... Learning unit, 51 ... Dialysis machine, 52 ... Blood inlet, 53 ... blood pump, 54 ... arterial drip chamber, 55 ... blood outlet, 56 ... intravenous drip chamber, 57 ... dialysate inlet, 58 ... dialysate pump, 59 ... dialysate outlet, R1 ... region, R2 ... region, 13a ... Blood pressure prediction model, 24a ... Learning data, 24b ... Blood pressure prediction model, 10a ... CPU, 10b ... Storage device, 10c ... Communication unit, 10d ... Input unit, 10e ... Output unit,

Claims (6)

An acquisition unit that acquires dialysis condition data for a specific patient,
The blood pressure prediction model generated based on machine learning using the dialysis condition data of each patient and the blood pressure value of each patient measured when performing dialysis treatment using the dialysis condition data is acquired. A prediction unit that inputs the dialysis condition data and generates a prediction value of the blood pressure of the specific patient.
A notification unit that notifies warning information about the specific patient based on the predicted value, and
A dialysis support device equipped with. In the machine learning, in addition to the dialysis condition data and the blood pressure value, the medical care data of each patient is used.
The acquisition unit acquires medical data of the specific patient and obtains medical data.
The predicted value is generated by inputting the dialysis condition data and the medical care data into the blood pressure prediction model.
The dialysis support device according to claim 1. When the predicted value is equal to or less than a predetermined threshold value or smaller than the predetermined threshold value, the notification unit notifies the warning information.
The dialysis support device according to claim 1 or 2. The notification unit notifies the predicted value, which is generated at a second time interval shorter than the first time interval for actually measuring blood pressure when performing dialysis treatment for the specific patient.
The dialysis support device according to any one of claims 1 to 3. The process of acquiring dialysis condition data for a specific patient,
The blood pressure prediction model generated based on machine learning using the dialysis condition data of each patient and the blood pressure value of each patient measured when performing dialysis treatment using the dialysis condition data is acquired. The step of inputting the dialysis condition data to generate the predicted value of the blood pressure of the specific patient, and
A step of notifying warning information about the specific patient based on the predicted value, and
Dialysis support method having. The arithmetic unit provided in the dialysis support device,
Acquisition department that acquires dialysis condition data for a specific patient,
The blood pressure prediction model generated based on machine learning using the dialysis condition data of each patient and the blood pressure value of each patient measured when performing dialysis treatment using the dialysis condition data is acquired. A prediction unit that inputs the dialysis condition data and generates a prediction value of the blood pressure of the specific patient.
A notification unit that notifies warning information about the specific patient based on the predicted value.
A dialysis support program that functions as a dialysis support program.

Images