Information processing device, information processing method, and program

Abstract

To easily estimate a maintenance timing for maintenance object components of a blood purification device, and to optimize the frequency and timing of the maintenance.SOLUTION: An information processing device for estimating a maintenance timing for a blood purification device for purifying a patient's blood includes: an operation data acquisition unit for acquiring operation data generated from state parameters indicating an operation state of the blood purification device; an estimation unit for inputting the operation data to a learned estimation model that has performed machine learning for estimating a maintenance timing for the blood purification device, and acquiring estimation data related to next maintenance of the blood purification device; and a control unit for generating display data for displaying maintenance information on the blood purification device on the basis of the estimation data. The learned estimation model is generated by the machine learning of the execution data on the past maintenance of the blood purification device.SELECTED DRAWING: Figure 7

Description

Technical Field

[0001] The present disclosure relates to an information processing apparatus, an information processing method, and a program for estimating the maintenance timing of a blood purification device.

Background Art

[0002] Conventionally, information processing using artificial intelligence (AI) has been performed, and research and development for applying the artificial intelligence to various technical fields have been underway. In recent years, the use cases of artificial intelligence have been increasing also in the medical field, and various judgment processes by artificial intelligence have become possible in place of doctors and other medical professionals.

[0003] For example, Patent Documents 1 to 3 describe the use of artificial intelligence in the medical field of blood purification. In particular, Patent Document 1 discloses a dialysis system in which artificial intelligence optimally adapts the state of an operation module of a medical device based on input information. Further, Patent Document 2 discloses a method of irradiating a patient's body fluid with light of a predetermined wavelength and predicting the patient's infectious disease by artificial intelligence from the optical characteristics of the body fluid obtained from a sensor. Furthermore, Patent Document 3 discloses a water filtration system that calculates the required quantity of consumable parts used in peritoneal dialysis and automatically places an order.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in a blood purification device, in addition to consumable parts that are replaced only once, there are maintenance target parts that require regular maintenance such as maintenance and management of their states. For such maintenance target parts, the frequency and timing of performing maintenance on each of them are set, so the maintenance of the blood purification device itself has become complicated.

[0006] The present disclosure has been made in view of such problems, and an object thereof is to provide an information processing device, an information processing method, and a program for easily estimating the maintenance timing of maintenance target parts of a blood purification device and optimizing the frequency and timing of maintenance.

Means for Solving the Problems

[0007] According to one aspect of the present disclosure, there is provided "an information processing device for estimating the maintenance timing of a blood purification device that purifies a patient's blood, the information processing device including: an operation data acquisition unit that acquires operation data generated from state parameters indicating the operation state of the blood purification device; an estimation unit that inputs the operation data into a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification device, and acquires estimation data related to the next maintenance of the blood purification device; and a control unit that generates display data for displaying the maintenance information of the blood purification device based on the estimation data, wherein the learned estimation model is generated by performing machine learning on the past maintenance execution data of the blood purification device."

[0008] According to one aspect of the present disclosure, there is provided an information processing method for estimating the maintenance timing of a blood purification device that purifies a patient's blood, the method including: obtaining operation data generated from state parameters indicating the operation state of the blood purification device; inputting the operation data into a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification device to obtain estimation data related to the next maintenance of the blood purification device; and generating display data for displaying maintenance information of the blood purification device based on the estimation data, wherein the learned estimation model is generated by performing machine learning on past maintenance execution data of the blood purification device.

[0009] According to one aspect of the present disclosure, there is provided a program for estimating the maintenance timing of a blood purification device that purifies a patient's blood, the program causing a computer to execute a process of obtaining operation data generated from state parameters indicating the operation state of the blood purification device, inputting the operation data into a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification device to obtain estimation data related to the next maintenance of the blood purification device, and generating display data for displaying maintenance information of the blood purification device based on the estimation data, wherein the learned estimation model is generated by performing machine learning on past maintenance execution data of the blood purification device.

Advantages of the Invention

[0010] According to the present disclosure, it is possible to provide an information processing apparatus, an information processing method, and a program for easily estimating the maintenance timing of maintenance target parts of a blood purification device and optimizing the frequency and timing of maintenance.

[0011] Note that the above effects are merely exemplary for convenience of explanation, and the effects of the present disclosure are not limited to the above. In addition to the above effects, according to the present disclosure, any effect described in the present disclosure can be achieved.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Embodiments for Carrying Out the Invention

[0013] Hereinafter, with reference to the drawings, the information processing apparatus of the present disclosure and the blood purification apparatus having the same will be described in detail. Note that the present invention is not limited to the content described below, and can be arbitrarily changed and implemented without changing the gist thereof. In addition, the drawings used in each embodiment schematically show the information processing apparatus according to the present disclosure and the blood purification apparatus having the same, and partial emphasis, enlargement, reduction, or omission, etc. are performed to deepen the understanding, and the scale, shape, etc. of each component may not accurately represent them. Furthermore, some numerical values used in each embodiment are all examples, and can be changed variously as needed. And, the same reference numerals are assigned to the common configurations in the drawings.

[0014] <First Embodiment> (Configuration of Blood Purification Apparatus) First, while referring to FIGS. 1 to 6, the configuration of the blood purification device having the information processing device of the present disclosure will be described. FIG. 1 is a schematic diagram showing an example of the usage state of the blood purification device according to the present embodiment. FIG. 2 is a block diagram showing the electrical configuration of the blood purification device according to the present embodiment. FIG. 3 is a configuration diagram of the internal piping section of the blood purification device according to the present embodiment. FIG. 4 is a configuration diagram of the extracorporeal circulation section of the blood purification device according to the present embodiment. FIGS. 5 and 6 are data tables stored in the blood purification device according to the present embodiment.

[0015] As shown in FIG. 1, the blood purification device 1 is composed of a dialysis device for performing dialysis treatment on the patient H. Specifically, the blood purification device 1 has a main body 3 installed on the base unit 2, a display 4 connected to the upper part of the main body 3, and a blood purifier 5 installed on the side of the main body 3. Further, the main body 3 of the blood purification device 1 has an information processing device 6, an internal piping section 7 for circulating dialysis fluid between the blood purifier 5, and an extracorporeal circulation section 8 for circulating the blood, which is the body fluid of the patient H, outside the body. With such a configuration, the blood purification device 1 can take out the blood of the patient H outside the body, remove unnecessary or toxic substances or moisture from the blood in the blood purifier 5, and return the purified blood to the patient H.

[0016] The base unit 2 is composed of a plate-shaped base 2a connected to the bottom of the main body 3 and four casters 2b installed on the base 2a. This makes it possible to easily move the blood purification device 1. Note that the number of casters 2b is not limited to four, and may be three or five or more as long as the blood purification device 1 can be moved.

[0017] The main body 3 is composed of a substantially rectangular parallelepiped housing. Also, various component devices and various components that make up the blood purification device 1 are arranged inside and on the surface of the main body 3. Here, as an example of the component devices, there are pumps such as a compound pump, a water removal pump, a degassing pump, a pressurizing pump, a stock solution pump, and a blood pump, which will be described later, and various detectors. These pumps, which are component devices, have various maintenance target parts such as valves, seal members, sliders, bearings, or impellers. In other words, pumps generally correspond to maintenance target parts. On the other hand, as an example of the components, there are parts such as solenoid valves, filters, various sensors, and blood circuits, which will be described later. These components belong to either maintenance target parts that require predetermined maintenance or consumable parts (for example, blood circuits, etc.) that need to be replaced by one hemodialysis treatment.

[0018] As shown in FIG. 2, in the blood purification device 1, a display 4, an information processing device 6, an internal piping section 7, an extracorporeal circulation section 8, and a communication section 9 are electrically connected to each other via a control line and a data line. Thereby, in the blood purification device 1, it is possible to transmit and receive signals, data, and information between various electrical components, and it is also possible to perform various controls by the information processing device 6. In the following, data is basically assumed to be composed of numerical values, symbols, characters, etc. obtained by processing signals and the like. Also, information is basically assumed to be what is obtained by collecting or processing the data. For example, it is assumed that the content can be used by the recipient as a material for subsequent consideration or can be utilized by the recipient. However, data and information may be used in a manner that does not conform to the above assumptions depending on their content and the context before and after.

[0019] In the present embodiment, the case of a hemodialysis device is described as an example of the blood purification device 1, but it is not limited thereto. For example, a device for acute blood purification, a peritoneal dialysis device, an ultrafiltration device, or a hemofiltration device can also be an example of the blood purification device 1. That is, in the present disclosure, any device may be used as long as the blood purification device 1 has maintenance target parts that require predetermined maintenance.

[0020] Note that, depending on the type of the blood purification device 1, the above-described constituent devices and components are different. Naturally, if the constituent devices and components are different, the maintenance target parts as the blood purification device 1 are also different. That is, the constituent devices, components, maintenance target devices, and maintenance target parts described in the present embodiment are merely examples, and even other devices and parts of the blood purification device 1 that correspond to the constituent devices and components and require replacement, maintenance, inspection, and servicing are applicable to the maintenance target devices and maintenance target parts. In other words, the maintenance target devices and maintenance target parts can be appropriately set by the administrator side of the blood purification device 1.

[0021] 〔Display〕 Next, as shown in FIGS. 1 and 2, the display 4 has an input unit 4a composed of a touch panel type input interface and an output unit 4b composed of a general screen type output interface. That is, the display 4 in the present embodiment is a touch panel provided with an input / output interface. Here, the detection method of input by the touch panel may be any method such as a capacitance type or a resistive film type.

[0022] Note that the input interface may be separated from the display 4. In this case, a keyboard provided with physical key buttons such as numeric keys or character input keys, and an input device such as a mouse may be provided in the blood purification device 1.

[0023] 〔Communication unit〕 Next, as shown in FIG. 2, the communication unit 9 is composed of a communication processing circuit 9a and an antenna 9b. The communication unit 9 transmits and receives information to and from a server device of a hospital installed separately from the blood purification device 1 or a terminal device used by a medical staff (administrator of the blood purification device 1) via the communication processing circuit 9a and the antenna 9b.

[0024] The communication processing circuit 9a may execute processing based on a broadband wireless communication method typified by the LTE method. Further, the communication processing circuit 9a may execute processing based on a method related to a narrowband wireless communication such as a wireless LAN typified by IEEE802.11 or Bluetooth (registered trademark). Furthermore, the communication processing circuit 9a may execute processing based on a method related to contactless wireless communication. And instead of or in addition to such wireless communication, wired communication may be used in the communication processing circuit 9a.

[0025] 〔Internal piping section〕 Next, as shown in FIG. 3, in the internal piping section 7, a main pipe L1 is connected to the liquid supply side of the blood purifier 5, and a main pipe L2 is connected to the liquid discharge side. Bypass pipes L3, L4, and L5 are connected between the main pipe L1 and the main pipe L2 so as to bypass the blood purifier 5. Further, a bypass pipe L10 is provided so as to connect the supply side (upstream side) of the main pipe L1 to the discharge side (downstream side) of the main pipe L2. Furthermore, bypass pipes L11 and L12 are connected in parallel to the main pipe L2 so as to bypass a part of the main pipe L2 in the middle of the main pipe L2. And a connecting pipe L13 is provided so as to connect the main pipe L2 and the bypass pipe L11. For example, a flexible material such as a vinyl chloride tube or a silicone tube is used for each pipe.

[0026] As shown in Fig. 3, pumps, valves, sensors, filters, etc. are arranged in each pipe. Specifically, in the main pipe L1, a pressure reducing valve V1, a solenoid valve V2, a degassing pump P0, a degassing chamber 10, a compound pump P1, a back pressure valve V3, a temperature sensor S1, a filter F1, a solenoid valve V4, a filter F2, a pressure sensor S2, and a solenoid valve V5 are arranged in order from the liquid supply terminal side of the internal pipe section 7 toward one end of the blood purifier 5. Further, a connector C1 is arranged at the tip of the main pipe L1 (the connection side with the blood purifier 5). Furthermore, in the main pipe L2, a solenoid valve V6, a pressure sensor S3, a pressurizing pump P2, a degassing chamber 11, a compound pump P1, a back pressure valve V7, a flow rate detector 12, and a solenoid valve V8 are arranged in order from one end of the blood purifier 5 toward the drain terminal side of the internal pipe section 7. And a connector C2 is arranged at the tip of the main pipe L2 (the connection side with the blood purifier 5).

[0027] The bypass pipe L3 is provided at a position connecting the filter F1 on the main pipe L1 and the downstream side of the pressure sensor S3 on the main pipe L2. Here, a solenoid valve V9 is arranged in the bypass pipe L3. Also, the bypass pipe L4 is provided at a position connecting the filter F2 on the main pipe L1 and the upstream side of the pressure sensor S3 on the main pipe L2 (that is, the downstream side of the solenoid valve V6). Here, a solenoid valve V10 is arranged in the bypass pipe L4. Furthermore, the bypass pipe L5 is provided at a position connecting the downstream side of the solenoid valve V5 on the main pipe L1 and the upstream side of the solenoid valve V6 on the main pipe L2. Here, the bypass pipe L5 is formed by connecting the connector C1 of the main pipe L1 and the connector C1 of the main pipe L2. That is, when performing treatment, the main pipe L1 and the main pipe L2 are connected to the blood purifier 5, but when cleaning the internal pipe section 7, etc., the connector C1 and the connector C2 are connected and the bypass connector C3 is formed on the bypass pipe L5. Note that the upstream and downstream are defined corresponding to the flow of the dialysate in each member.

[0028] The bypass pipe L10 is connected via a degassing chamber 10 between the solenoid valve V2 and the degassing pump P0 on the main pipe L1 and the downstream side of the solenoid valve V8 on the main pipe L2, and has a configuration that bypasses components other than the pressure reducing valve V1, the solenoid valve V2, and the degassing chamber 10. The bypass pipes L11, L12 are connected between the pressure sensor S3 and the compound pump P1 on the main pipe L2 and between the compound pump P1 and the flow rate detector 12 on the main pipe L2, and have a configuration that bypasses the compound pump P1. In the present embodiment, the bypass pipe L11 is connected closer to the compound pump P1 than the bypass pipe L12. Further, a back pressure valve V11 and a water removal pump P3 are provided in the bypass pipe L11. Furthermore, a solenoid valve V12 is provided in the bypass pipe L12. The connection pipe L13 connects the degassing chamber 11 on the main pipe L2 and between the back pressure valve V11 and the water removal pump P3 in the bypass pipe L11.

[0029] Each of the various valves, various pumps, and the flow rate detector 12 described above has its operations controlled based on a control signal supplied from the information processing device 6. Thereby, the dialysate can be circulated at a desired flow rate or various types of cleaning can be performed. For example, in order to more accurately perform the liquid feeding amount by the compound pump P1, when the compound pump P1 is performing a suction operation, control is performed to support the suction operation by pressurizing with a water supply pressure or the pressurizing pump P2. That is, the pressure reducing valve V1 is controlled to adjust the water supply pressure for supporting the suction amount on the liquid supply side, and the pressurizing pump P2 is controlled for supporting the suction amount on the drainage side. Here, the pressure control of the pressurizing pump P2 is executed when the pressure of the pressurizing pump P2 becomes equal to or higher than a predetermined value and the back pressure valve V11 opens. On the other hand, when the compound pump P1 is performing a discharge operation, when the discharge pressure becomes equal to or lower than a predetermined value, the back pressure valves V3, V7 are closed so that the liquid does not flow, and control is performed to support the discharge operation by eliminating the influence of inertia. That is, for supporting the discharge amount on the liquid supply side, the back pressure valve V3 is closed, and for supporting the discharge amount on the drainage side, the back pressure valve V7 is closed.

[0030] In addition, these valves, pumps, and the flow rate detector 12 can transmit state parameters indicating the operating state of each component device or component part, or the blood purification device 1 having these, to the information processing device 6. For example, these valves, pumps, and the flow rate detector 12 may transmit the state parameters measured by various built-in sensors to the information processing device 6. That is, the state parameters may be measured values measured by various sensors. In the case where various sensors are not built in, the state parameters may be measured and transmitted by various sensors arranged in the vicinity of each component device or component part. And the state parameters measured in the internal piping section 7 are collectively referred to as piping section parameters.

[0031] Similarly, the various sensors described above can also transmit the measured values to the information processing device 6. Thereby, the information processing device 6 can acquire various state parameters indicating the operating state of the internal piping section 7 of the blood purification device 1. On the other hand, regarding the various filters, O-rings, connectors, and ports described above, since they do not have a sensing function, the usage status of each may be grasped by a temperature sensor or a pressure sensor provided in the vicinity thereof.

[0032] As described above, in the internal piping section 7, in addition to the circulation process and the cleaning process of the dialysate, the acquisition process of the state parameters indicating the operating state of the blood purification device 1 and the component devices and component parts of the blood purification device 1 is performed. As will be described later, the acquired state parameters will be used for estimating the maintenance timing of the blood purification device 1.

[0033] 〔Extracorporeal circulation section〕 Next, as shown in FIG. 4, the extracorporeal circulation unit 8 has a structure in which an arterial blood circuit L21 is connected to the blood introduction side of the blood purifier 5, and a venous blood circuit L22 is connected to the blood discharge side of the blood purifier 5. Further, a liquid level adjustment circuit L23 is connected between the arterial blood circuit L21 and the venous blood circuit L22 so as to bypass the blood purifier 5. Here, the liquid level adjustment circuit L23 is composed of a bypass line L24 arranged in parallel with the blood purifier 5 and an open line L25 for introducing or discharging air to or from the bypass line L24. For example, flexible materials such as vinyl chloride tubes or silicone tubes are used for each circuit.

[0034] In the arterial blood circuit L21, a connector C21, an electromagnetic valve V21, a bubble detector 22, a blood pump P21, a blood concentration detector 23, and an arterial air trap chamber 24 are arranged in order from the patient H side toward the blood purifier 5. In the venous blood circuit L22, a venous air trap chamber 25, a flow rate detector 26, a pressure detector 27, a bubble detector 28, an electromagnetic valve V22, and a connector C22 are arranged in order from the blood purifier 5 toward the patient H. When the blood purification device 1 is driven to purify the blood of the patient H, an arterial puncture needle (not shown) is connected to the connector C21, and a venous puncture needle (not shown) is connected to the connector C22, and each puncture needle is inserted into the arm of the patient H.

[0035] With such a configuration, in the arterial blood circuit L21, the amount and concentration of the blood taken out from the patient H are detected, and in the venous blood circuit L22, the amount and concentration of the blood returned to the patient H are detected.

[0036] In the bypass line L24 of the liquid level adjustment circuit L23, a pressure sensor S22, a solenoid valve V23, a solenoid valve V24, and a pressure sensor S23 are arranged in order from the arterial side air trap chamber 24 toward the venous side air trap chamber 25. Further, a liquid level adjustment pump P22 is arranged in the open line L25 of the liquid level adjustment circuit L23. With such a configuration, it is possible to drive the liquid level adjustment pump to introduce or discharge air, and the blood level in each air trap chamber can be adjusted.

[0037] Each of the solenoid valves, various pumps, and various detectors described above is controlled in each operation based on a control signal supplied from the information processing device 6. Further, these solenoid valves, pumps, and detectors can transmit state parameters indicating the operating state of each component device or component, or the blood purification device 1 having them, to the information processing device 6. For example, these solenoid valves, pumps, and detectors may transmit state parameters measured by various built-in sensors to the information processing device 6. That is, the state parameter may be a measured value measured by various sensors. When various sensors are not built in, the state parameters may be measured and transmitted by various sensors arranged in the vicinity of each component device or component. And, the state parameters measured in the extracorporeal circulation unit 8 are collectively referred to as circulation unit parameters.

[0038] Similarly, each of the pressure sensors described above can also transmit the measured value to the information processing device 6. Thereby, the information processing device 6 can acquire various state parameters indicating the operating state of the extracorporeal circulation unit 8 of the blood purification device 1. On the other hand, since each of the connectors, filters, and air trap chambers described above does not have a sensing function, the usage status of each may be grasped by a temperature sensor or a pressure sensor provided in the vicinity thereof.

[0039] As described above, in the extracorporeal circulation unit 8, in addition to the processes of introducing and discharging blood and adjusting the liquid level, the blood purification device 1 and the acquisition process of state parameters indicating the operating states of the component devices and components of the blood purification device 1 are performed. As will be described later, the acquired state parameters will be used for estimating the maintenance timing of the blood purification device 1.

[0040] 〔Information Processing Device〕 Next, as shown in FIG. 2, the information processing device 6 according to the present embodiment is composed of a processor 6a and a memory 6b.

[0041] The processor 6a is composed of a GPU (Graphics Processing Unit) or a CPU (Central Processing Unit), and functions as a control unit that controls other connected component devices or components based on various programs stored in the memory 6b. Specifically, the processor 6a reads and executes a program for executing blood purification processing or a program for executing an OS from the memory 6b. In addition, the processor 6a executes a process for estimating the maintenance timing of the blood purification device 1 and the component devices and components of the blood purification device 1. Note that the processor 6a may be composed of a single GPU or CPU, or may be composed of a combination of multiple CPUs or GPUs.

[0042] The memory 6b is composed of a ROM, a RAM, a non-volatile memory, an HDD, etc., and functions as a storage unit. The ROM stores an instruction command for executing blood purification processing as a program. In addition, the ROM stores a learned estimation model necessary for estimating the maintenance timing of the blood purification device 1 and the component devices and components of the blood purification device 1. The RAM is used for writing and reading data while the program stored in the ROM is being processed by the processor 6a. The non-volatile memory is a storage device in which writing and reading of data are executed by the execution of the program, and the data written here is saved even after the execution of the program is completed.

[0043] In particular, in the present embodiment, a program for estimating the maintenance timing of the blood purification device 1 for purifying a patient's blood is stored. The program causes a computer (i.e., the information processing device 6 of the blood purification device 1) to execute a process of acquiring operation data generated from state parameters indicating the operation state of the blood purification device 1. Further, the program causes the computer to execute a process of inputting the operation data into a learned estimation model that has performed machine learning for estimating the maintenance timing of the blood purification device 1 and acquiring estimation data related to the next maintenance of the blood purification device 1. Furthermore, the program causes the computer to execute a process of generating display data for displaying the maintenance information of the blood purification device 1 based on the estimation data. Note that the processes executed by the program will be described later.

[0044] Also, in the memory 6b, the operation states of the blood purification device 1, its constituent devices, and its constituent components as shown in FIG. 5 are stored. As shown in FIG. 5, as an example of the constituent devices and constituent components, they constitute the internal piping section 7 and the extracorporeal circulation section 8 and are arranged in each pipe or each circuit. As an example of storing the operation states, in the data table of FIG. 5, for the double pump, the water removal pump, the degassing pump, the pressurizing pump, the first electromagnetic valve (corresponding to the electromagnetic valve V2), the second electromagnetic valve (corresponding to the electromagnetic valve V4), the first filter (corresponding to the filter F1), and the second filter (corresponding to the filter F2), the "operation times (driving times)", "temperature", "pressure", "flow rate", "usage time", "usage times", "high temperature time", "washing times", and "treatment times" at the current time are stored as the operation states.

[0045] More specifically, for each pump and solenoid valve, "number of operations", "temperature", "pressure", "flow rate", "usage time", "number of uses", and "high-temperature time" are stored as operating states. On the other hand, the "number of operations" is not stored for the filter, and "temperature", "pressure", "flow rate", "usage time", "number of uses", "high-temperature time", "number of cleaning times", and "number of treatment times" are stored as operating states. For the hemodialysis device, "usage time", "number of uses", "number of cleaning times", and "number of treatment times" are stored as operating states.

[0046] Note that for the data table shown in FIG. 5, only some of the constituent devices and components are described. Naturally, the operating states of other constituent devices and components shown in FIGS. 3 and 4 may also be stored. Also, the types of operating states are not limited to those shown in the table of FIG. 5, and any data indicating various operations of the blood purification device 1, its constituent devices, and its components can be stored in the data table. For example, an operating state such as "low-temperature time" may be added.

[0047] In addition, the memory 6b stores maintenance information on the blood purification device 1, its constituent devices, and its components as shown in FIG. 6. As shown in FIG. 6, an example of the constituent devices and components includes those that constitute the internal piping section 7 and the extracorporeal circulation section 8 and are disposed in each pipe or each circuit. Specifically, the data table stores information regarding "constituent device / component", "component name", "operation / replacement", "previous maintenance date and time", and "order source" for the duplex pump, water removal pump, degassing pump, pressure pump, first solenoid valve (corresponding to solenoid valve V2), second solenoid valve (corresponding to solenoid valve V4), first filter (corresponding to filter F1), and second filter (corresponding to filter F2). Here, "operation / replacement" is information regarding the timing that serves as a guide from the state of a new product or the state after maintenance to the next maintenance.

[0048] More specifically, for each pump, the maintenance target parts are stored in the "Part Name" column, and the "Operation / Replacement", "Previous Maintenance Date and Time", and "Supplier" of each part are stored. On the other hand, for the solenoid valve and the filter, since they themselves are maintenance target parts, the target part name is not stored in the "Part Name" column, and the information related to "Operation / Replacement", "Previous Maintenance Date and Time", and "Supplier" is stored. Also, as "Operation / Replacement", a predetermined time is stored for each component device and component part, but instead of time, the number of uses, the date and time of use, or the usage amount may be stored, or further, it may be data combining these.

[0049] Note that for the data table shown in FIG. 6, only some of the component devices and component parts are described. Naturally, maintenance information regarding other component devices and component parts shown in FIGS. 3 and 4 may also be stored. Also, the types of maintenance information are not limited to those shown in the table of FIG. 6, and as long as it is content related to the maintenance of the blood purification device 1, its component devices, and its component parts, it can be stored in the data table. For example, maintenance information such as "amount" may be added.

[0050] (Functional Configuration of Information Processing Device) Next, while referring to FIG. 7, the functional configuration of the information processing device 6 of the blood purification device 1 according to the present embodiment will be described. Here, FIG. 7 is a functional block diagram of the information processing device 6 according to the first embodiment. In particular, in FIG. 7, in addition to the information processing device 6, other component devices of the blood purification device 1 are also described, and the flow of information and data between each device is also described.

[0051] As shown in FIG. 7, the information processing apparatus 6 includes an operation data acquisition unit 51, an estimation unit 52, a control unit 53, and a storage unit 54. The operation data acquisition unit 51 includes a calculation unit 55. Further, the estimation unit 52 includes a learned estimation model 56. Each of these units is realized by the functions of the processor 6a and the memory 6b of the information processing apparatus 6 itself, or by the processor 6a reading and executing a program stored in the memory 6b.

[0052] The operation data acquisition unit 51 receives various signals, which are state parameters indicating the operation state of the blood purification apparatus 1, from the pump P61, the pressure sensor S62, the temperature sensor S63, and the flow rate sensor S64. Here, the pump P61 is a general term for various pumps shown in FIGS. 3 and 4, and is not limited to any one pump. Similarly, the pressure sensor S62 is a general term for each pressure sensor shown in FIGS. 3 and 4 and the pressure sensors included in other constituent devices (for example, the pressure detector 27) other than the pump, and is not limited to any one pressure sensor. Also, the temperature sensor S63 is a general term for each temperature sensor shown in FIGS. 3 and 4 and the temperature sensors included in other constituent devices (for example, various valves) other than the pump, and is not limited to any one temperature sensor. Further, the flow rate sensor S64 is a general term for the flow rate sensors included in each flow rate detector shown in FIGS. 3 and 4 and the flow rate sensors included in other constituent devices (for example, various valves), and is not limited to any one flow rate sensor.

[0053] The operation data acquisition unit 51 receives an operation signal, a temperature signal, a pressure signal, and a flow rate signal from the pump P61. The operation signal is an electrical signal related to the operation of the pump P61 itself, for example, an electrical signal related to the total number of rotations of the impeller or the number of driving times of the diaphragm. The temperature signal, the pressure signal, and the flow rate signal are electrical signals detected by each sensor provided in the pump P61, and are electrical signals related to the temperature, pressure, and flow rate of the dialysate or blood inside the pump P61. Note that the temperature signal of the pump P61 may be an electrical signal related to the temperature of the pump P61 itself.

[0054] In addition, the operation data acquisition unit 51 also receives input data from the input unit 4a as state parameters indicating the operation state of the blood purification device 1. For example, the input data is the operation time, operation frequency, or treatment frequency of the blood purification device 1 input by the administrator of the blood purification device 1. Note that the input data is not necessarily received and is positioned as additional data when generating the operation data described later.

[0055] The calculation unit 55 of the operation data acquisition unit 51 calculates the operation state from the received various electrical signals. Specifically, the calculation unit 55 calculates the current operation state of the pump P61 based on the various electrical signals received from the pump P61. Examples of the operation state include "usage time", "usage frequency", and "high temperature time" as shown in FIG. 5, but operation states other than these may also be calculated. For example, the calculation unit 55 may calculate the usage time and usage frequency while determining whether the pump P61 is operating based on the operation signal, pressure signal, and flow rate signal. Further, the calculation unit 55 may calculate the high temperature time of the pump P61 based on the operation signal and temperature signal.

[0056] In addition, the calculation unit 55 calculates the current operation state of the component device or component part equipped with each sensor, or the current operation state of the component device or component part located near each sensor, based on the various electrical signals received from the pressure sensor S62, temperature sensor S63, and flow rate sensor S64. Examples of the operation state include "operation frequency", "usage time", "usage frequency", and "high temperature time" as shown in FIG. 5, but operation states other than these may also be calculated. For example, the calculation unit 55 may calculate the operation frequency, usage time, and usage frequency while determining whether the solenoid valve is operating based on the pressure signal and flow rate signal. Similarly, the calculation unit 55 may calculate the high temperature time of the solenoid valve based on the calculated operation frequency and the received temperature signal. Similarly, the calculation unit 55 may calculate the usage time, usage frequency, and high temperature time of the filter based on the pressure signal, flow rate signal, and temperature signal.

[0057] Furthermore, the calculation unit 55 calculates the operating state of the blood purification device 1 based on the operating data of the above-described constituent devices and components. Examples of the operating state include "usage time", "usage frequency", "washing frequency", and "treatment frequency" as shown in FIG. 5, but operating states other than these may also be calculated. For example, the calculation unit 55 may calculate "usage time", "usage frequency", "washing frequency", and "treatment frequency" while determining the current operating state (stopped, pre-treatment preparation, in-treatment, post-treatment washing) of the blood purification device 1 from the operating states of the constituent devices and components. Note that the calculation unit 55 may calculate "usage time", "usage frequency", "washing frequency", and "treatment frequency" based on the input data received from the input unit 4a, or may set them as "usage time", "usage frequency", "washing frequency", and "treatment frequency" without performing an operation on the received input data.

[0058] In the present embodiment, when the calculation of the operating state by the calculation unit 55 is completed, the operation data acquisition unit 51 aggregates each operating state into one piece of operation data. That is, the operation data acquisition unit 51 acquires operation data for input to the learned estimation model 56 from the calculated operating states of the blood purification device 1, the constituent devices, and the components. In other words, the operation data includes data related to the operating states of the blood purification device 1, the constituent devices, and the components. Note that the calculation of the operating state itself may be executed outside the information processing device 6 using each constituent device or another calculator or the like. In this case, the operation data acquisition unit 51 simply receives various operating states from the outside without having the calculation unit 55. Further, the operation data does not necessarily include the above-described plurality of operating states. For example, one piece of operation data may be configured by one operating state. In this case, the operating state is used as the operation data as it is.

[0059] The operation data acquisition unit 51 transmits the acquired operation data to the storage unit 54 in order to store the operation data. Further, the operation data acquisition unit 51 may also transmit various signals that are the source of the operation data to the storage unit 54. The storage unit 54 stores the received operation data in the memory 6b. Furthermore, the operation data acquisition unit 51 transmits the acquired operation data to the estimation unit 52 in order to estimate the maintenance timing of the blood purification device 1.

[0060] The estimation unit 52 receives maintenance data from the storage unit 54 in addition to the received operation data in order to acquire estimation data using the learned estimation model 56. Here, the maintenance data is the "previous maintenance date and time" for each constituent device and component shown in the data table of FIG. 6. Further, the estimation unit 52 inputs the received operation data and maintenance data into the learned estimation model 56, and acquires estimation data related to the next maintenance of the blood purification device 1. Here, the estimation data includes data on the maintenance timing of the blood purification device 1 and the parts to be maintained. Note that the learned estimation model 56 and the estimation data will be described later including the model settings. When the estimation unit 52 acquires the estimation data that is the output result of the learned estimation model 56, it transmits the estimation data to the control unit 53.

[0061] The control unit 53 receives order destination data from the storage unit 54 in addition to the received estimation data in order to notify the administrator of the blood purification device 1 of the maintenance information of the blood purification device 1. Here, the order destination data is the "order destination" for each constituent device and component shown in the data table of FIG. 6. The control unit 53 generates display data and a control signal serving as a display instruction for display based on the estimation data and the order destination data, and transmits the display instruction (control signal) together with the display data to the output unit 4b. Further, the control unit 53 may select the order destination of the parts to be maintained included in the estimation data from the order destination data and automatically place an order for the parts to be maintained. In this case, information indicating that the parts to be maintained have been ordered may be added to the display data.

[0062] The output unit 4b displays the maintenance information of the blood purification device 1 via the display 4, and notifies the administrator of the blood purification device 1 of the necessity for maintenance. By checking the maintenance information, the administrator will appropriately perform the maintenance related to the blood purification device 1.

[0063] (Processing in the blood purification device) Next, before estimating and displaying the maintenance information of the blood purification device 1 according to the present embodiment, the processing executed in the blood purification device 1 will be described with reference to FIGS. 8 to 11. Here, FIG. 8 is a sequence diagram related to the estimation of maintenance in the blood purification device 1 of the present embodiment. FIG. 9 is a flowchart showing the flow of setting the learned estimation model according to the present embodiment. FIG. 10 is a schematic diagram showing the flow of evaluation of the learned estimation model set in the information processing device 6 of the present embodiment. FIG. 11 is an example of notification of maintenance information displayed in the blood purification device 1 according to the present embodiment.

[0064] First, as shown in FIG. 8, in the information processing device 6, initial settings are made (S111). Here, the initial settings refer to the preparatory processing necessary for estimating the maintenance information of the blood purification device 1 using artificial intelligence (AI). That is, through the initial settings, a learned estimation model 56 is generated, and the learned estimation model 56 is implemented in the information processing device 6.

[0065] Specifically, as shown in FIG. 9, the processor 6a of the information processing device 6 acquires learning data for generating the learned estimation model 56 (S201). Here, as an example of the learning data, as shown in FIG. 10, usage status data of the blood purification device 1, past failure data, maintenance execution data, and basic maintenance data may be used. These data may be input by the administrator of the blood purification device 1 via the input unit 4a, or may be received via the communication unit 9.

[0066] The usage data may include, for example, the usage time, number of uses, number of treatments, and number of cleanings of the blood purification device 1. Further, the usage data may include, for example, the number of operations, usage time, and high-temperature time of the constituent devices and components of the blood purification device 1. Furthermore, the usage data may include, for example, data related to the temperature, pressure, and flow rate detected by the blood purification device 1 and the constituent devices and components of the blood purification device 1. That is, the usage data may be composed of data related to the operating status of the blood purification device 1 and the constituent devices and components of the blood purification device 1, calculated based on the data detected from the blood purification device 1.

[0067] The past failure data may include, for example, the name of the failed component, the date and time of the failure, and data related to the usage status at the time of the failure. Specifically, in the case of the pump P61, in addition to the names of components such as the valve, impeller, and seal, the date and time of the failure and data related to the usage status of the pump up to the date and time of the failure may be used as the past failure data. In this case, it is preferable that the past failure data is linked to the above-described usage data.

[0068] The maintenance implementation data may include, for example, the name of the device or component to be maintained, the date and time of maintenance, and data related to the maintenance content. Specifically, when the impeller fails in the pump P61, in addition to the name of the pump P61 (for example, double pump, dewatering pump, degassing pump, pressurizing pump, blood pump, liquid level adjustment pump, etc.), the details of the impeller, the impeller replacement work, and data related to the replacement timing may be used as the data related to the maintenance content for the maintenance implementation data. In this case, it is preferable that the maintenance implementation data is linked to the above-described past failure data.

[0069] The basic maintenance data may include, for example, the name of the component device or component to be maintained in the blood purification device 1, the name of the component to be maintained, and data related to the replacement time. Here, the replacement time is synonymous with "operation / replacement" in FIG. 6, and is the normal durability time from a new or maintained state to the next replacement. Specifically, in the case of the pump P61, the name of the pump P61 (such as a duplex pump, a water removal pump, a degassing pump, a pressurizing pump, a blood pump, a liquid level adjustment pump, etc.), the name of the component to be maintained (such as an impeller, a valve, a poppet valve, etc.), and the replacement time of each component may be used as the basic maintenance data. Note that, similar to "operation / replacement" in FIG. 6, the number of uses, the use date and time, or the usage amount may be used instead of the replacement time, and further, data combining these may be used.

[0070] As described above, in the present embodiment, in order to generate a learned estimation model, the past usage status data, past failure data, past maintenance implementation data, and basic maintenance data of the blood purification device 1 are associated and used. However, although it is preferable to associate all of these data, some data may be excluded, or additional data may be added, etc., and the association between the data may be appropriately changed according to the machine learning described later.

[0071] Next, the processor 6a of the information processing device 6 performs annotation processing on the acquired learning data (S202). As a specific annotation processing, annotations or the like are attached to the acquired learning data to generate teacher data. For example, the processor 6a of the information processing device 6 tags the past maintenance date and time for each component device and component, and generates correct answer data in which the maintenance timing corresponding to a predetermined operating state is associated for each component device and component.

[0072] Next, the processor 6a of the information processing apparatus 6 performs machine learning using the acquired learning data and the annotated teacher data (S203). As an example, in this machine learning, the learning data and the teacher data are given to a neural network composed of a combination of neurons, and learning is repeated while adjusting the parameters of each neuron so that the output of the neural network becomes the same as the correct data among the teacher data. Note that the above machine learning is merely an example, and machine learning using scoring may be performed.

[0073] Next, the processor 6a of the information processing apparatus 6 evaluates the learned estimation model 56 generated by the machine learning (S204). Here, the processor 6a of the information processing apparatus 6 uses evaluation data different from the learning data used in the machine learning and executes the model evaluation. For example, as shown in FIG. 10, the processor 6a of the information processing apparatus 6 inputs data related to the usage time, the number of operations, the temperature, the pressure, the flow rate, the number of treatments, the number of washings, and the past maintenance of the blood purification apparatus 1 and the constituent devices and components of the blood purification apparatus 1. Then, the processor 6a of the information processing apparatus 6 evaluates whether the estimation data output from the learned estimation model 56 is a correct result. Here, as the estimation data, for example, data related to the maintenance timing, the target device, and the target component is output. As a specific evaluation method, the processor 6a of the information processing apparatus 6 determines whether the actual maintenance information corresponding to the evaluation data matches the estimation data. Here, the actual maintenance information is information related to the maintenance actually performed next to the past maintenance input as the evaluation data. Note that if the actual maintenance information does not match the estimation data, the process starts over from the acquisition of the learning data, and machine learning is performed again.

[0074] Next, when the actual maintenance information corresponding to the evaluation data matches the estimated data, the processor 6a of the information processing apparatus 6 implements the generated learned estimation model 56 (S205). Specifically, the processor 6a of the information processing apparatus 6 stores the generated learned estimation model 56 in the memory 6b. As a result, the estimation unit 52 having the learned estimation model 56 functions in the information processing apparatus 6.

[0075] Returning to FIG. 8, after the initial setting (S111) is completed, data acquisition in the internal piping unit 7 is performed (S112). Specifically, measurement of the piping unit parameters among the state parameters indicating the operating state of the blood purification apparatus 1 is performed by various constituent devices and various sensors provided in the internal piping unit 7. Then, the piping unit parameters acquired in the internal piping unit 7 are transmitted to the information processing apparatus 6 (T111).

[0076] Also, after the initial setting (S111) is completed, data acquisition in the extracorporeal circulation unit 8 is performed (S113). Specifically, measurement of the circulation unit parameters among the state parameters indicating the operating state of the blood purification apparatus 1 is performed by various constituent devices and various sensors provided in the extracorporeal circulation unit 8. Then, the circulation unit parameters acquired in the extracorporeal circulation unit 8 are transmitted to the information processing apparatus 6 (T112).

[0077] Furthermore, after the initial setting (S111) is completed, data input in the display 4 is performed (S114). Specifically, corresponding to an input operation by the administrator of the blood purification apparatus 1, data indicating the operating state of the blood purification apparatus 1 is input via the input unit 4a of the display 4. Then, the input data input from the display 4 is transmitted to the information processing apparatus 6 (T113).

[0078] Next, in the information processing apparatus 6, calculation processing of operation data is performed (S115). Specifically, when the operation data acquisition unit 51 of the information processing apparatus 6 receives the pipe unit parameters, the circulation unit parameters, and the input data, the calculation unit 55 of the operation data acquisition unit 51 generates the operation data of the blood purification apparatus 1 for input to the learned estimation model based on the received parameters and data. Thereby, the process of acquiring the operation data by the operation data acquisition unit 51 of the information processing apparatus 6 (operation data acquisition process) is completed.

[0079] Next, in the information processing apparatus 6, storage processing of the received pipe unit parameters, circulation unit parameters, input data, and the operation data calculated in S115 is performed (S116). Specifically, the operation data acquisition unit of the information processing apparatus 6 transmits these parameters and data to the storage unit 54, and the storage unit 54 stores these parameters and data in the memory 6b.

[0080] Next, in the information processing apparatus 6, acquisition processing of estimation data related to the next maintenance of the blood purification apparatus 1 is performed using the calculated operation data (S117). Specifically, the operation data acquisition unit 51 of the information processing apparatus 6 transmits the acquired operation data to the estimation unit 52. Also, the estimation unit 52 acquires past maintenance data from the storage unit 54. Thereafter, the estimation unit 52 inputs the operation data and the maintenance data to the learned estimation model 56. The learned estimation model 56 performs estimation processing based on the input data and outputs the estimation data as the estimation result. Thereby, the process in which the estimation unit 52 acquires the estimation data related to the next maintenance of the blood purification apparatus 1 and outputs the estimation data is completed. Note that, as an example of the estimation data, data related to the maintenance timing, the target apparatus, and the target parts as shown in FIG. 10 is included. For this reason, the state change of the blood purification apparatus 1 from the time of past maintenance implementation to the time of operation data acquisition is associated with the said estimation data.

[0081] Next, in the information processing apparatus 6, a process for generating display data for notifying the administrator of the blood purification apparatus 1 of the maintenance information of the blood purification apparatus 1 is performed (S118). Specifically, the estimation unit 52 of the information processing apparatus 6 transmits the acquired estimated data to the control unit 53. The control unit 53 acquires order destination data related to the maintenance target apparatus and the maintenance target parts from the storage unit 54. Thereafter, the control unit 53 performs an ordering process for the maintenance target parts based on the acquired two pieces of data, and generates display data including a maintenance report for displaying the maintenance information and a control signal for causing the display data to be displayed. Thereby, the process of generating display data for displaying the maintenance information of the blood purification apparatus 1 based on the estimated data is completed.

[0082] Next, a transmission process of the display data and the control signal is performed from the information processing apparatus 6 to the display 4 (T114). Specifically, the control unit 53 of the information processing apparatus 6 transmits a control signal that is a display instruction to the output unit 4b together with the display data to be displayed on the output unit 4b of the display 4. Thereafter, in the display 4, a maintenance notification based on the received display data is performed by the output unit 4b (S119).

[0083] Here, FIG. 11 shows an example of an image output by the output unit 4b. As shown in FIG. 11, two display units 61 and 62 are formed in the output unit 4b. A sentence (notification of the next maintenance) for notifying that maintenance is required within a predetermined period is displayed on the display unit 61. On the other hand, the maintenance time, the maintenance target apparatus, and the maintenance target parts, which are the estimation results of the learned estimation model 56, are displayed on the display unit 62. Further, a button 63 is displayed on the display unit 62. The button 63 is a button for displaying an explanation about how to perform the required maintenance. Thereby, when the administrator of the blood purification apparatus 1 touches the button 63, an explanation of the maintenance is separately displayed on the output unit 4b.

[0084] (Processing in the information processing apparatus) Next, while referring to FIG. 12, a series of processing flows executed in the information processing apparatus 6 in the process of estimating maintenance described in FIG. 8 will be specifically described. Here, FIG. 12 is a flowchart related to the estimation of maintenance executed in the information processing apparatus 6 according to the present embodiment. Specifically, FIG. 12 is a flowchart related to the processing and determination from S115 to S118 in FIG. 8. Note that the processing flow is mainly performed by the processor 6a of the information processing apparatus 6 reading and executing a program stored in the memory 6b.

[0085] First, the processor 6a of the information processing apparatus 6 determines whether data related to state parameters indicating the operating state of the blood purification apparatus 1 has been received from the internal piping unit 7, the extracorporeal circulation unit 8, and the input unit 4a (S301). Here, if the processor 6a of the information processing apparatus 6 does not receive the data (S301: No), this step is repeated and the process does not proceed to the next step.

[0086] On the other hand, when the processor 6a of the information processing apparatus 6 receives the data (S301: Yes), the processor 6a of the information processing apparatus 6 determines whether calculation based on the received data is necessary (S302). Specifically, the processor 6a of the information processing apparatus 6 functions the operation data acquisition unit 51 and determines whether the data received by the operation data acquisition unit 51 can be directly used as operation data and whether all the data necessary for generating operation data is available. For example, when the received data is an operation signal, a temperature signal, a pressure signal, or a flow rate signal, it is necessary to calculate the number of operations, the numerical value of the temperature, the value of the pressure, or the value of the flow rate from these electrical signals. Also, even when the received data is the number of operations, the numerical value of the temperature, the value of the pressure, or the value of the flow rate, if the usage time, the number of uses, the high-temperature time, the number of washings, and the number of treatments can be calculated, it is determined that calculation is necessary. In other words, the processor 6a of the information processing apparatus 6 determines whether the data to be included in the operation data can be calculated based on the received data.

[0087] If it is determined in S302 that the calculation is unnecessary (S302: No), the next step, S303, is skipped and the process proceeds to S304. On the other hand, if it is determined in S303 that the calculation is necessary (S302: Yes), the processor 6a of the information processing apparatus 6 causes the calculation unit 55 of the operation data acquisition unit 51 to function and executes the calculation of the data to be included in the operation data (S303). Specifically, as described above, the number of operations, temperature values, pressure values, and flow rate values are calculated from various electrical signals, and further, the usage time, number of uses, high-temperature time, number of cleaning times, and number of treatment times are calculated.

[0088] Next, the processor 6a of the information processing apparatus 6 generates operation data using the received data or calculation results (S304). Specifically, the processor 6a of the information processing apparatus 6 causes the operation data acquisition unit 51 to function, combines the number of operations, temperature values, pressure values, flow rate values, usage time, number of uses, high-temperature time, number of cleaning times, and number of treatment times into one, and generates operation data for input to the learned estimation model 56.

[0089] Next, the processor 6a of the information processing apparatus 6 performs a process of storing the generated operation data and various received signals in the memory 6b (S305). Specifically, the processor 6a of the information processing apparatus 6 causes the operation data acquisition unit 51 and the storage unit 54 to function, transmits the operation data and the signal from the operation data acquisition unit 51 to the storage unit 54, and the operation data and the signal are stored in the memory 6b by the storage unit 54.

[0090] Next, the processor 6a of the information processing apparatus 6 performs a process of estimating the next maintenance of the blood purification apparatus 1 (S306). Specifically, the processor 6a of the information processing apparatus 6 functions as the operation data acquisition unit 51, the estimation unit 52, the storage unit 54, and the learned estimation model 56. The operation data acquisition unit 51 transmits operation data to the estimation unit 52, and the storage unit 54 transmits maintenance data to the estimation unit 52. Further, the estimation unit 52 inputs the received operation data and maintenance data into the learned estimation model 56. Subsequently, the learned estimation model 56 uses the input data to perform an estimation process related to the next maintenance of the blood purification apparatus 1. Thereafter, the learned estimation model 56 outputs estimation data that is the result of the estimation process related to the next maintenance, and the estimation unit 52 acquires the estimation data.

[0091] Next, the processor 6a of the information processing apparatus 6 determines whether maintenance of the blood purification apparatus 1 is necessary within a predetermined period based on the acquired estimation data (S307). Here, the predetermined period is, for example, a period within several weeks or within several months, and may be determined in consideration of ordering of parts to be maintained. Specifically, the processor 6a of the information processing apparatus 6 functions as the estimation unit 52 and the control unit 53. The estimation unit 52 transmits the estimation data to the control unit 53. Thereafter, the control unit 53 determines whether the maintenance time included in the estimation data is within the set predetermined period.

[0092] If it is determined in S307 that maintenance is not required within a predetermined period (S307: No), this flow ends. On the other hand, if it is determined in S307 that maintenance is required within a predetermined period (S307: Yes), the processor 6a of the information processing apparatus 6 determines whether it is necessary to order parts (S308). Specifically, the processor 6a of the information processing apparatus 6 causes the control unit 53 to function, and the control unit 53 determines whether the parts to be maintained are parts to be ordered according to the content of maintenance included in the estimated data. For example, if the maintenance content is part replacement, the control unit 53 determines that it is necessary to order parts, and if the maintenance content is operation confirmation of the constituent devices or constituent parts, the control unit 53 determines that it is not necessary to order parts.

[0093] If it is determined in S308 that it is not necessary to order parts (S308: No), the next step, S309, is skipped and the process proceeds to S310. On the other hand, if it is determined in S308 that it is necessary to order parts (S308: Yes), the processor 6a of the information processing apparatus 6 performs an ordering process for the parts to be maintained (S309). Specifically, the processor 6a of the information processing apparatus 6 causes the control unit 53 and the storage unit 54 to function, and the storage unit 54 transmits ordering destination data to the control unit 53. Subsequently, the control unit 53 selects an ordering destination for the parts to be maintained included in the estimated data from the received ordering destination data. After that, the control unit 53 places an order for the parts to be maintained with the selected ordering destination. For example, the control unit 53 may read out an ordering template stored in the memory 6b to create ordering request data, and transmit the created ordering request data to the ordering destination via the communication unit.

[0094] Next, the processor 6a of the information processing apparatus 6 generates display data for notifying maintenance (S310). Specifically, the processor 6a of the information processing apparatus 6 causes the control unit 53 to function, and based on the estimated data, the control unit 53 generates display data including information on the maintenance timing, the maintenance target apparatus, and the maintenance target parts of the blood purification apparatus 1, and a control signal for displaying the display data. Further, the control unit 53 may add information related to parts ordering to the display data. That is, when parts have been ordered, the control unit 53 may add information for indicating that fact to the display data.

[0095] (Modification of the First Embodiment) In the above embodiment, as shown in FIG. 10, the learned estimation model 56 output the maintenance timing of the blood purification apparatus 1 as estimated data, but the output content of the estimated data may be made more detailed. For example, as shown in FIG. 13, the learned estimation model 56 may estimate the maintenance target parts and the maintenance timing for each component device of the blood purification apparatus 1 and output the estimation result. Here, FIG. 13 is a schematic diagram related to a modification of the estimation of maintenance executed in the information processing apparatus 6 according to the present embodiment, shown in the same manner as FIG. 10. By performing such an estimation of the maintenance timing, it becomes possible to manage the maintenance of the blood purification apparatus 1 in a subdivided manner.

[0096] In such a case, the control unit 53 generates display data and control signals for displaying maintenance target parts and maintenance times for each component device of the blood purification device 1 as maintenance information. For this reason, a display as shown in FIG. 14 may be made on the output unit 4b of the display 4. Here, FIG. 14 is a modified example of notification of maintenance information displayed in the blood purification device 1 according to the first embodiment. As shown in FIG. 14, in the output unit 4b, in addition to the two display units 61 and 62, a display unit 64 is added. In the added display unit 64, a button 65 for displaying a maintenance explanation is provided in the same manner as in the display unit 62. Further, in the display unit 62 and the display unit 64, the maintenance time and the maintenance target parts are displayed for each maintenance target device. In particular, even when there are a plurality of maintenance target parts, the display is made in one display unit 62.

[0097] Also, a scroll bar 66 is displayed adjacent to the display unit 62 and the display unit 64. When there are a plurality of maintenance target parts and they cannot be displayed within one screen, by moving the scroll bar 66 downward, the display unit related to other maintenance target parts will appear. Thereby, the administrator of the blood purification device 1 can check the detailed information of the maintenance target parts one by one.

[0098] In the above embodiment, supervised learning using teacher data has been performed for the generation of the learned estimation model 56, but it is not limited to this. For example, an unsupervised learning or reinforcement learning generally used may be utilized to generate the learned estimation model 56. That is, as long as the maintenance time of the blood purification device 1 can be estimated, the method of machine learning is not limited. Naturally, if the method of machine learning is different, since the learning data is different, the learning data for generating the learned estimation model 56 is not limited to the past usage status data, past failure data, past maintenance implementation data, and basic maintenance data of the blood purification device 1, and various data related to the blood purification device 1 can be used.

[0099] In the above-described embodiment, a specific date and time was estimated as the maintenance timing of the blood purification device 1, but the maintenance timing may be estimated as a period with a certain margin for the timing. That is, the learned estimation model 56 may estimate the maintenance periods of the blood purification device 1, the component devices, and the components.

[0100] In the above-described embodiment, the ordering of parts required for maintenance was automatically executed by the control unit 53, but a list of parts to be ordered may be displayed on the output unit 4b of the display 4 by the processing of the control unit 53, and the ordering of parts may be executed by the button operation of the display 4 by the administrator of the blood purification device 1. Further, the part ordering by the administrator and the automatic part ordering according to the above-described embodiment may be preset, and the administrator of the blood purification device 1 may select any of the ordering processes.

[0101] In the above-described embodiment, a piping configuration that enables the introduction and derivation of dialysis fluid using the dual pump P1 was adopted. However, for example, two diaphragm pumps may be used instead of the dual pump P1. An example in such a case will be described as a modification with reference to FIG. 15. Here, FIG. 15 is a configuration diagram of a blood purification device 31 according to a modification of the present embodiment. Note that the same devices and components as those of the blood purification device 1 are denoted by the same reference numerals, and the description thereof will be omitted.

[0102] As shown in FIG. 15, the blood purification device 31 according to the modification has an internal piping section 37 for introducing and discharging dialysis fluid to and from the blood purifier 5, and an extracorporeal circulation section 38 for introducing and discharging blood to and from the blood purifier 5. Note that the internal piping section 37 and the internal piping section 7 are functionally the same, but their constituent members and the like are different. Similarly, the extracorporeal circulation section 38 and the extracorporeal circulation section 8 are functionally the same, but their constituent members and the like are different.

[0103] The internal piping section 37 has a structure in which the main pipe L31 is connected to the liquid supply side of the blood purifier 5 and the main pipe L32 is connected to the liquid discharge side. In the main pipe L31, a pressure reducing valve V1, a solenoid valve V2, a dialysate adjusting section 91, a diaphragm pump P31, a solenoid valve V31, and a connector C1 are arranged in order from the liquid supply terminal side of the internal piping section 37 toward one end of the blood purifier 5. Also, in the main pipe L32, a solenoid valve V32, a diaphragm pump P32, and a flow rate detector 12 are arranged in order from one end of the blood purifier 5 toward the liquid discharge terminal side of the internal piping section 37. In the internal piping section 37, it is possible to perform water removal by adjusting the flow rates of the two diaphragm pumps P31 and P32. For this reason, a water removal pump P3 is not arranged in the internal piping section 37.

[0104] The extracorporeal circulation section 38 has a structure in which an arterial side blood circuit L33 is connected to the blood introduction side with respect to the blood purifier 5 and a venous side blood circuit L34 is connected to the blood derivation side with respect to the blood purifier 5. In the arterial side blood circuit L33, a connector C21, an arterial side clamp CL31, and a blood pump P21 are arranged in order from the patient H side toward the blood purifier 5. Also, in the venous side blood circuit L34, a venous side air trap chamber 25, a flow rate detector 26, a bubble detector 28, a venous side clamp CL32, and a connector C22 are arranged in order from the blood purifier 5 toward the patient H. Further, a supply pipe L35 for supplying a priming liquid is connected between the arterial side clamp CL31 and the blood pump P21. And a supply side clamp CL33 is arranged in the supply pipe L35, and the supply pipe L35 is connected to a storage bag 92. A physiological saline solution as a priming liquid is stored in the storage bag 92.

[0105] As shown in FIG. 15, the internal piping section 37 and the extracorporeal circulation section 38 are connected by a communication pipe L41. Specifically, one end of the communication pipe L41 (on the side of the internal piping section 37) is connected between the solenoid valve V32 and the diaphragm pump P32 of the internal piping section 37, and the other end (on the side of the extracorporeal circulation section 38) is connected to the venous side air trap chamber 25. Also, a solenoid valve V33 is provided in the communication pipe L41 as a component of the internal piping section 37.

[0106] In such a modification, the diaphragm pumps P31 and P32 and the solenoid valves V31 to V32 correspond to the parts to be maintained, and the maintenance timing and the like are estimated in the same manner as in the above-described embodiment.

[0107] In the blood purification device 1 in the above-described embodiment, a compound pump system in which the introduction and discharge of the dialysate are performed using the compound pump P1 is employed, but the present invention is not limited thereto. For example, two chambers partitioned into two chambers by a single diaphragm may be installed, the amount of the dialysate and the drainage amount may be controlled to be equal, and a double chamber system in which dehydration is controlled by a dehydration pump may be employed.

[0108] As an example of adopting the double chamber system, in the internal piping section, a pressure reducing valve, a first solenoid valve, a supply side diaphragm pump, a flow rate adjusting valve, a flow meter, a concentration sensor, a temperature sensor, and a second solenoid valve are arranged in order from the supply liquid side in the supply side piping for introducing the dialysate to the blood purifier. Also, a third solenoid valve, a dialysate pressure sensor, a filtrate pump, and a discharge side diaphragm pump are arranged in order from the blood purifier side in the discharge side piping for discharging the dialysate from the blood purifier. Further, a dehydration pump and a negative pressure circulation pump are arranged in another pipe branched from the discharge side pipe.

[0109] As another example of adopting the double-chamber method, in the internal piping section, a pump for introducing two types of chemical solutions toward the blood purifier is provided in the supply-side piping for introducing dialysate into the blood purifier. Also, in the supply-side piping, a balancing chamber valve is provided near each of the inlets of the diaphragm pumps. On the other hand, a balancing chamber valve, a dialysate filter, a holding valve, a water removal pump, and a discharge valve are arranged in the discharge-side piping for discharging dialysate from the blood purifier. Further, a heat exchanger, a balancing chamber valve, a flow pump, an air separation pump, a load pressure valve, and a recirculation valve are arranged in the circulation piping provided by branching from the discharge-side piping.

[0110] In a modification example where the double-chamber method is adopted, each component arranged in each pipe corresponds to a maintenance target component, and the estimation of the maintenance timing and the like is performed in the same manner as in the above embodiment.

[0111] Furthermore, for example, a viscus chamber method may be adopted in which the inside of the blocked viscus chamber is partitioned into three chambers by two diaphragms, the volume of the central viscus chamber is changed, and a difference in volume is generated between the dialysate chamber and the drainage chamber to control water removal. That is, in the viscus chamber method, a water removal pump is not used.

[0112] When the viscus chamber method is adopted, in the internal piping section, the viscus chamber of the viscus chamber partitioned into three chambers is connected via a viscus pump, and the introduction and discharge of viscus oil (silicone oil) are possible. Also, a three-way solenoid valve, a two-way solenoid valve, a temperature sensor, and a pressure sensor are arranged in the pipe provided from one viscus chamber toward the blood purifier. Further, a hydraulic pump and a relief valve are arranged in the pipe provided from the blood purifier toward the other viscus chamber.

[0113] In a modification that adopts the biscuit chamber method, each component arranged in each pipe corresponds to a maintenance target component, and the maintenance timing and the like are estimated in the same manner as in the above embodiment.

[0114] (Operation and Effect of the First Embodiment) In this embodiment, when the current operation data of the blood purification device 1 is input to the learned estimation model 56 generated by associating and machine learning the past usage status data, past failure data, past maintenance implementation data, and basic maintenance data of the blood purification device 1, the learned estimation model 56 estimates the maintenance timing of the blood purification device 1. That is, the learned estimation model 56 estimates the maintenance timing of the maintenance target devices and target components that are maintenance targets among the constituent devices and components that make up the blood purification device 1. Therefore, it is not necessary for the administrator of the blood purification device 1 to estimate the maintenance timing, and it becomes easy and highly accurate for the administrator to obtain maintenance information including the maintenance timing. Further, since the learned estimation model 56 is generated by associating and machine learning the various data described above, the frequency and timing of maintenance related to the estimation result can be optimized.

[0115] <Second Embodiment> In the first embodiment, one maintenance timing was estimated for each blood purification device or device to be maintained (constituent device). However, a range may be provided for the maintenance timing for each device or component to be maintained, and further, the maintenance timing for the devices or components to be maintained may be unified. That is, when estimating the maintenance period of a blood purification device, a device to be maintained, or a component to be maintained, an estimation may be made that takes into account the maintenance periods of different maintenance targets and sets the overlapping periods as the maintenance period. Such a case will be described as the second embodiment with reference to FIGS. 16 and 17. Here, FIG. 16 is a schematic diagram related to the estimation of maintenance executed in the information processing device 6 according to the present embodiment. FIG. 17 is an example of notification of maintenance information displayed on the blood purification device 1 according to the present embodiment. Note that parts different from the first embodiment will be basically described, and the description of the same content will be omitted, and the same reference numerals will be used in the drawings.

[0116] As shown in FIG. 16, in the present embodiment, maintenance unification data is added as learning data. Here, the maintenance unification data is data in which an allowable range is set for the maintenance timing for each device or component to be maintained. The maintenance unification data is associated with the basic maintenance data and allows for a margin in the replacement time of the basic maintenance data. For example, for each of the constituent devices and components shown in FIG. 6, an allowable range for maintenance may be set in units of months, weeks, dates, or hours. Note that in FIG. 16, it is assumed that the allowable ranges for the poppet valve, cap seal, and coupling, which are constituent parts of the pump, are set in terms of dates as the maintenance unification data.

[0117] The unified maintenance data described above is added as learning data, and a learned estimation model 156 is generated by performing machine learning in the same manner as the flow shown in FIG. 9 of the first embodiment. As shown in FIG. 16, the learned estimation model 156 according to the present embodiment estimates the maintenance period of the maintenance target parts for each constituent device and includes the maintenance period in the estimation data and outputs it even when the same input data as in the first embodiment is input.

[0118] In addition, when there are a plurality of maintenance target parts, the learned estimation model 156 estimates the maintenance period so as to align the maintenance at the same timing. For example, when the maintenance period of the poppet valve of the duplex pump is estimated to be from December 10, 2023 to December 31, 2023, and the maintenance period of the coupling of the drainage pump is estimated to be from November 20, 2023 to December 15, 2023, the learned estimation model 156 estimates the maintenance periods of the poppet valve and the coupling to be from December 10, 2023 to December 15, 2023. That is, the learned estimation model 156 outputs the common maintenance period of each maintenance target part as the final estimation result. In other words, the learned estimation model 156 estimates so as to perform the maintenance performed in a predetermined period at the same timing. As a result, it is possible to reduce the number of maintenance times and extend the usage period of the blood purification device 1.

[0119] Furthermore, the learned estimation model 156 estimates suitable usage conditions for delaying the maintenance timing as much as possible during the estimated maintenance period. For example, if the maintenance period of the poppet valve of a duplex pump is estimated to be from December 10, 2023 to December 31, 2023, and the maintenance period of the coupling of the drainage pump is estimated to be from November 20, 2023 to December 15, 2023, in order to output the common period from December 10, 2023 to December 15, 2023 as the maintenance period, it is necessary to delay the maintenance period of the coupling of the drainage pump as much as possible. Therefore, the learned estimation model 156 includes conditions for delaying the maintenance period in the estimation data as suitable usage conditions and outputs them. The suitable usage conditions are such that they limit the operating conditions of each device or component to be maintained. More specifically, conditions such as the flow rate and temperature of the pump used are included in the operating conditions. This makes it possible to reduce the number of maintenance times and extend the usage period of the blood purification device 1.

[0120] Note that in order to enable the above output by the learned estimation model 156, it is preferable to appropriately perform annotation processing based on the above-described learning data to generate teacher data for enabling desired output.

[0121] Based on the estimated data with the limited maintenance period as described above and further including suitable usage conditions, maintenance information is notified from the output unit 4b. For example, as shown in FIG. 17, in the display unit 62 and the display unit 64, a predetermined period is displayed in the item of the maintenance timing. Further, the display unit 71 is also displayed in the output unit 4b. In the display unit 71, the suitable usage conditions included in the estimated data are displayed, and a button 72 for confirming a detailed explanation thereof is displayed. Thereby, when the administrator of the blood purification device 1 touches the button 72, the suitable usage conditions of the blood purification device 1, the device to be maintained, or the component to be maintained are separately displayed.

[0122] In this embodiment, the learned estimation model 156 estimated that the maintenance times were at the same timing. However, the learned estimation model 156 may output the maintenance periods of the respective parts to be maintained as estimation data, and the control unit 53 may unify the maintenance times based on the estimation data. In other words, the unification of the maintenance times may be processed without relying on AI.

[0123] (Operation and Effect of the Second Embodiment) In this embodiment, machine learning is performed by adding the maintenance unification data as learning data to generate the learned estimation model 156. The learned estimation model 156 outputs estimation data so as to align the maintenance times of the devices and parts to be maintained. In addition, the learned estimation model 156 outputs the estimation data including the suitable operating conditions for aligning the maintenance times. Therefore, it is possible to reduce the number of maintenance times and increase the usage period of the blood purification device 1. Furthermore, since the suitable operating conditions of the blood purification device 1 are notified to the administrator, the administrator can easily determine the usage schedule and the like of the blood purification device 1.

[0124] <Third Embodiment> In the second embodiment, as conditions for delaying the maintenance time, the suitable operating conditions of the blood purification device 1, the device to be maintained, or the part to be maintained were included in the estimation data. Instead of this, conditions regarding the selection of patients may be included. That is, a proposal for the next patient to be treated by the blood purification device 1 may be made. Such a case will be described as the third embodiment with reference to FIGS. 18 and 19. Here, FIG. 18 is a schematic diagram related to the estimation of maintenance executed in the information processing apparatus 6 according to this embodiment. FIG. 19 is an example of notification of maintenance information displayed in the blood purification device 1 according to this embodiment. Note that parts different from the first and second embodiments will be basically described, and descriptions of the same content will be omitted, and the same reference numerals will be used in the drawings.

[0125] As shown in FIG. 18, in this embodiment, patient blood purification performance data is added as learning data. Here, the patient blood purification performance data is the conditions of blood purification treatment set for each patient H. The unified maintenance data is associated with the basic maintenance data and the unified maintenance data. The conditions of the blood purification treatment may include, for example, the amount of dialysate, dialysis time, operating conditions of each device, and the like.

[0126] By adding the above-described patient blood purification performance data as learning data and performing machine learning in the same manner as the flow shown in FIG. 9 of the first embodiment, a learned estimation model 256 is generated. As shown in FIG. 18, the learned estimation model 256 according to this embodiment estimates the maintenance period of the maintenance target parts for each constituent device even when the same input data as in the first embodiment is input, and also estimates the next patient who can preferably use the blood purification device 1, and outputs the data related to the next patient included in the estimated data.

[0127] In order to enable the above output by the learned estimation model 256, it is preferable to appropriately perform annotation processing based on the above-described learning data to generate teacher data for performing a desired output.

[0128] Based on the estimated data including the limited maintenance period and the preferable use conditions as described above, the output unit 4b notifies the maintenance information. For example, as shown in FIG. 19, in the display unit 62 and the display unit 64, a predetermined period is displayed in the item of the maintenance time. Further, similar to the second embodiment, the display unit 71 is also displayed on the output unit 4b. The display unit 71 displays the conditions of the next patient, which are the preferable use conditions included in the estimated data, and a button 72 for confirming a detailed explanation thereof. Thereby, when the administrator of the blood purification device 1 touches the button 72, more detailed data of the proposed patient H is separately displayed as the conditions for preferably using the blood purification device 1.

[0129] Using the estimated data in which the next patient is selected as described above, the output unit 4b notifies maintenance information. For example, as shown in FIG. 19, on the display unit 71, as suitable usage conditions included in the estimated data, a specific patient is described, and a button 72 for confirming a detailed explanation thereof is displayed. Thus, when the administrator of the blood purification device 1 touches the button 82, it becomes possible to confirm the detailed data of the patient H who is recommended to be treated by the blood purification device 1.

[0130] (Operation and Effect of the Third Embodiment) In the present embodiment, machine learning is performed by adding the blood purification performance data of patients as learning data, and a learned estimation model 256 is generated. Then, the learned estimation model 256 outputs estimated data so as to align the maintenance times of the devices and parts to be maintained, and outputs the data of the next patient, which is preferable as suitable operating conditions for aligning the maintenance times, included in the estimated data. For this reason, it becomes possible to reduce the number of maintenance times, and the usage period of the blood purification device 1 can be increased. Furthermore, since the administrator of the blood purification device 1 can confirm the data of the next suitable patient, it is possible to easily determine the allocation of patients to the blood purification device 1 and the like.

[0131] <Fourth Embodiment> In the first to third embodiments, the estimation related to the next maintenance has been performed in the blood purification device 1, but the estimation process may be performed in a server device that manages a plurality of blood purification devices 1. Such a case will be described as the fourth embodiment with reference to FIGS. 20 to 23. Here, FIG. 20 is a schematic diagram showing the configuration of the blood purification system according to the present embodiment. FIG. 21 is a functional block diagram of the blood purification device according to the present embodiment. FIG. 22 is a block diagram showing the physical configuration of the information processing device according to the present embodiment. FIG. 23 is a functional block diagram of the blood purification device according to the present embodiment. Note that parts different from the first to third embodiments will be basically described, and the description of the same content will be omitted, and the same reference numerals will be given in the drawings.

[0132] First, as shown in FIG. 20, a blood purification system 300 according to the present embodiment includes a plurality of blood purification devices 1, a server device 400, and a mobile terminal device 500. Also, two blood purification devices 1, the server device 400, and the mobile terminal device 500 are communicably connected to each other by a network 600 such as the Internet. Note that the network 600 may be configured by wireless, wired, or a combination thereof.

[0133] In the blood purification system 300, each of the blood purification devices 1 transmits various data acquired within the device to the server device 400 or the mobile terminal device 500 via the network 600. Also, the server device 400 processes various data received from the blood purification device 1 and transmits the processing result to the blood purification device 1 or the mobile terminal device 500. Further, the mobile terminal device 500 can also process the received various data and transmit it to the blood purification device 1 or the server device.

[0134] The configuration of the blood purification device 1 of the present embodiment is basically the same as the configuration of the first embodiment. As a different part, since the operation data acquisition of the blood purification device 1, the acquisition of estimated data related to the next maintenance, and the processing related to the control based on the estimated data are not performed, the blood purification device 1 does not have the configuration and functions related to the processing. Note that in FIG. 20, only two blood purification devices 1 are shown, but the number of blood purification devices 1 is not limited to this, and the blood purification system 300 may have three or more blood purification devices 1. Also, the blood purification device 1 is not limited to the same type of device, and devices with different types and versions may be used.

[0135] The server device 400 is a computer installed in a hospital, manages various data in the hospital, and performs various data processing in the hospital. Note that the installation location of the server device 400 is not limited to a hospital and may be outside the hospital. Also, in FIG. 20, the server device 400 is shown as a single one, but it is also possible to distribute various configurations and processes of the server device 400 described later to a plurality of other server devices or cloud server devices.

[0136] The mobile terminal device 500 may be a wireless communication-enabled device typified by a smartphone, but is not limited thereto. For example, the mobile terminal device 500 may be a feature phone, a personal digital assistant, a PDA, a laptop personal computer, a desktop personal computer, or the like. In FIG. 20, only one mobile terminal device 500 is shown, but the number of mobile terminal devices 500 is not limited thereto, and the blood purification system 300 may have two or more mobile terminal devices 500.

[0137] Next, while referring to FIG. 21, the functional configuration of the blood purification device 1 according to the present embodiment will be described. In particular, in FIG. 21, the same reference numerals are given to the same devices, components, and configurations as in the first embodiment.

[0138] As shown in FIG. 21, the information processing device 6 has a calculation unit 55, which is realized by the processor 6a of the information processing device 6 reading and executing a program stored in the memory 6b. The calculation unit 55 receives various signals, which are state parameters indicating the operating state of the blood purification device 1, from the pump P61, the pressure sensor S62, the temperature sensor S63, and the flow rate sensor S64. Further, the calculation unit 55 also receives input data from the input unit 4a as a state parameter indicating the operating state of the blood purification device 1.

[0139] The calculation unit 55 calculates the operating state from the received various electrical signals. Specifically, the calculation unit 55 calculates the current operating state of the pump P61 based on the various electrical signals received from the pump P61. Further, the calculation unit 55 calculates the current operating state of the component device or component part equipped with each sensor, or calculates the current operating state of the component device or component part located in the vicinity of each sensor, based on the various electrical signals received from the pressure sensor S62, the temperature sensor S63, and the flow rate sensor S64. Furthermore, the calculation unit 55 calculates the operating state of the blood purification device 1 based on the operation data of the above-described component devices and component parts. Then, the calculation unit 55 summarizes each operating state into one piece of operation data. Here, the operation data includes data related to the operating states of the blood purification device 1, the component devices, and the component parts.

[0140] Thereafter, the calculation unit 55 transmits the generated one piece of operation data to the outside of the blood purification device 1 via the communication unit 9. In the present embodiment, the calculation unit 55 transmits the operation data to the server device 400.

[0141] Next, while referring to FIGS. 22 and 23, the configuration of the server device 400 according to the present embodiment will be described. First, as shown in FIG. 21, the server device 400 includes an input unit 404a, an output unit 404b, an information processing device 406, and a communication unit 409. The information processing device 406 includes a processor 406a and a memory 406b. Further, the communication unit 409 includes a communication processing circuit 409a and an antenna 409b.

[0142] The processor 406a of the information processing device 406 is composed of a GPU or a CPU, and functions as a control unit that performs various calculations and controls based on various programs stored in the memory 406b. Note that the processor 406a may be composed of a single GPU or CPU, or may be composed of a combination of a plurality of CPUs or GPUs.

[0143] The memory 406b is composed of a ROM, a RAM, a non-volatile memory, an HDD, etc., and functions as a storage unit. The ROM stores the learned estimation model necessary for estimating the maintenance timing of the blood purification device 1 and the constituent devices and components of the blood purification device 1. Also, the RAM is used for writing and reading data while the program stored in the ROM is being processed by the processor 406a. The non-volatile memory is a storage device in which writing and reading of data are executed by the execution of the program, and the data written therein is stored even after the execution of the program is completed.

[0144] In particular, in the present embodiment, a program for estimating the maintenance timing of the blood purification device 1 that purifies the patient's blood is stored. The program causes the server device 400, which is a computer, to execute a process of acquiring operation data generated from state parameters indicating the operation state of the blood purification device 1. Also, the program causes the computer to execute a process of inputting the operation data into a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification device 1 and acquiring estimation data related to the next maintenance of the blood purification device 1. Furthermore, the program causes the computer to execute a process of generating display data for displaying the maintenance information of the blood purification device 1 based on the estimation data. Note that the processes executed by the program will be described later.

[0145] Also, in the memory 406b, similar to the memory 6b of the information processing device 6 according to the first embodiment, the operation states of the blood purification device 1, its constituent devices, and its components are stored. Specifically, the operation states shown in FIG. 5 are stored. However, since the server device 400 manages a plurality of blood purification devices 1, each operation state is stored in the memory 406b so that each of the blood purification devices 1 can be identified.

[0146] Furthermore, the memory 406b stores maintenance information on the blood purification device 1, its constituent devices, and its constituent components, similar to the memory 6b of the information processing device 6 according to the first embodiment. Specifically, the maintenance information shown in FIG. 6 is stored. However, since the server device 400 manages a plurality of blood purification devices 1, each piece of maintenance information is stored in the memory 406b so that each of the blood purification devices 1 can be identified.

[0147] The communication unit 409 transmits and receives information to and from the blood purification device 1 or the mobile terminal device 500 installed separately from the server device 400 via the communication processing circuit 409a and the antenna 409b.

[0148] The communication processing circuit 409a may execute processing based on a broadband wireless communication method typified by the LTE method. Also, the communication processing circuit 409a may execute processing based on a method related to a wireless LAN typified by IEEE802.11 or a narrowband wireless communication such as Bluetooth (registered trademark). Furthermore, the communication processing circuit 409a may execute processing based on a method related to non-contact wireless communication. And instead of or in addition to such wireless communication, wired communication may be used for the communication processing circuit 409a.

[0149] Next, as shown in FIG. 23, the information processing device 406 includes an operation data acquisition unit 451, an estimation unit 452, a control unit 453, and a storage unit 454. Also, the estimation unit 452 includes a learned estimation model 456. Each of these units is realized by the processor 406a and the memory 406b of the information processing device 406 itself functioning, or by the processor 406a reading and executing a program stored in the memory 406b.

[0150] The operation data acquisition unit 451 acquires operation data transmitted from the blood purification device 1 via the communication unit 409. Further, the operation data acquisition unit 451 transmits the acquired operation data to the storage unit 454 in order to store the operation data. Further, the operation data acquisition unit 451 may also transmit various signals that are the source of the operation data to the storage unit 454. The storage unit 454 stores the received operation data in the memory 406b. Further, the operation data acquisition unit 451 transmits the acquired operation data to the estimation unit 452 in order to estimate the maintenance timing of the blood purification device 1 from the acquired operation data.

[0151] The estimation unit 452 receives maintenance data from the storage unit 454 in addition to the received operation data in order to acquire estimation data using the learned estimation model 456. Here, the maintenance data is the "previous maintenance date and time" for each constituent device and component shown in the data table of FIG. 6, similar to that according to the first embodiment. Further, the estimation unit 452 inputs the received operation data and maintenance data into the learned estimation model 456 and acquires estimation data related to the next maintenance of the blood purification device 1. Here, the estimation data includes data on the maintenance timing of the blood purification device 1 and the parts to be maintained.

[0152] The learned estimation model 456 according to the present embodiment is the same as the learned estimation model 56 according to the first embodiment and is generated by the same learning data and machine learning. Therefore, a detailed description thereof is omitted.

[0153] After that, when the estimation unit 452 acquires the estimated data that is the output result of the learned estimation model 456, the estimation unit 452 transmits the estimated data to the control unit 453. In order to notify the administrator of the blood purification device 1 of the maintenance information of the blood purification device 1, the control unit 453 receives order destination data from the storage unit 454 in addition to the received estimated data. Here, the order destination data is the "order destination" for each constituent device and constituent component shown in the data table of FIG. 6, similar to that according to the first embodiment. Based on the estimated data and the order destination data, the control unit 453 generates display data for display as maintenance information, and transmits the display data to the communication unit 409. Further, the control unit 453 may select the order destination of the maintenance target part included in the estimated data from the order destination data, and automatically place an order for the maintenance target part. In this case, information indicating that the maintenance target part has been ordered may be added to the display data.

[0154] The communication unit 409 transmits the received display data to the blood purification device 1 that has transmitted the operation data. The communication unit 409 also transmits the received display data to the mobile terminal device 500. As a result, the administrator of the blood purification device 1 can check the information related to the next maintenance of the blood purification device 1 displayed on the display 4 of the blood purification device 1 or the mobile terminal device 500 operated by the administrator himself / herself.

[0155] In the above description, the learned estimation model 456 was assumed to be the same as the learned estimation model 56 according to the first embodiment. However, naturally, the learned estimation model 456 may be the same as the learned estimation model 156 according to the second embodiment or the learned estimation model 256 according to the third embodiment. In particular, when the learned estimation model 456 is the same as the learned estimation model 256 according to the third embodiment, the learned estimation model 456 is generated by performing machine learning while associating the past usage status data, past failure data, past maintenance execution data, basic maintenance data, and blood purification performance data of a plurality of the patients of the blood purification device 1. Then, the learned estimation model 456 can estimate each of a plurality of blood purification devices by including the next patient selection data in the estimation data. As a result, for the entire hospital where a plurality of blood purification devices 1 are installed, the allocation of patients to the blood purification device 1 and the like can be easily determined, and the hospital operation related to blood purification treatment can be smoothly performed. That is, it becomes possible to grasp the status of the blood purification device 1 in the entire hospital and establish a maintenance schedule for the entire hospital. For example, when it is possible to classify the blood purification device 1 used under severe conditions where the pump speed is high and the blood purification device 1 used for relatively slow treatment from data on the treatment method, dialysate flow rate, or blood pump speed, the blood purification device 1 can be automatically allocated to each patient according to the patient's treatment conditions so that a plurality of blood purification devices 1 in the hospital can be used in a well-balanced manner. For example, while allocating the blood purification device 1 that has been used under severe conditions to a patient who undergoes the relatively slow treatment, the blood purification device 1 that has been used for the relatively slow treatment can be allocated to a patient who undergoes the treatment under severe conditions so that it can be used for the treatment under the relevant conditions.

[0156] In addition, in the above-described embodiment, the estimation related to the next maintenance is performed in the server device 400, but the present invention is not limited to this. For example, a learned estimation model may be implemented in the information processing device of the mobile terminal device 500 to execute the estimation process.

[0157] (Operation and Effect of the Fourth Embodiment) In this embodiment, when current operation data of the blood purification device 1 is input to a learned estimation model 456 generated by performing machine learning by associating past usage status data, past failure data, past maintenance implementation data, and basic maintenance data of the blood purification device 1, the learned estimation model 456 estimates the maintenance timing of the blood purification device 1. That is, the learned estimation model 456 estimates the maintenance timing of the maintenance target devices and target components that are the maintenance targets among the component devices and components constituting the blood purification device 1. For this reason, it is not necessary for the administrator of the blood purification device 1 to estimate the maintenance timing, and the administrator can easily and highly accurately obtain maintenance information including the maintenance timing. Further, since the learned estimation model 456 is generated by performing machine learning by associating the above-described various types of data, the frequency and timing of maintenance related to the estimation result can be optimized.

[0158] Also, in this embodiment, in a device different from the blood purification device 1, an estimation process is performed, and the results of the estimation process are centrally managed. For this reason, the maintenance timings of a plurality of blood purification devices 1 and information related thereto can be centrally managed, and it becomes possible to more easily perform maintenance and inspection of the plurality of blood purification devices 1 installed in a hospital.

[0159] <Embodiment of the Present Disclosure> The first embodiment of the present disclosure is an information processing device that estimates the maintenance timing of a blood purification device that purifies a patient's blood, the information processing device including: an operation data acquisition unit that acquires operation data generated from state parameters indicating the operation state of the blood purification device; an estimation unit that inputs the operation data to a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification device, and acquires estimation data related to the next maintenance of the blood purification device; and a control unit that generates display data for displaying maintenance information of the blood purification device based on the estimation data.

[0160] In this way, by simply inputting the operation data of the blood purification device into the learned estimation model, the maintenance timing of the maintenance target parts of the blood purification device is output. Therefore, information related to maintenance can be easily estimated, and the frequency and timing of the maintenance can be optimized.

[0161] According to a second embodiment of the present disclosure, in the first embodiment, the estimation data includes data of maintenance target parts of the blood purification device. Thereby, it becomes possible to grasp the specific parts to be maintained.

[0162] According to a third embodiment of the present disclosure, in the first or second embodiment, the learned estimation model is generated by performing machine learning while associating past usage status data, past failure data, past maintenance implementation data, and basic maintenance data of the blood purification device. Thereby, the estimation accuracy by the learned estimation model can be improved, and more reliable estimation data can be obtained.

[0163] According to a fourth embodiment of the present disclosure, in any one of the first to third embodiments, the state parameter includes at least one of an operation signal, a temperature signal, a pressure signal, or a flow rate signal detected by a constituent device constituting the blood purification device or a sensor provided in a flow path of the blood purification device. Thereby, it becomes possible to obtain estimation data corresponding to the current operation state of the blood purification device.

[0164] In the fifth embodiment of the present disclosure, in the fourth embodiment, the learned estimation model estimates maintenance target parts and maintenance timing for the maintenance target parts for each of the constituent devices, and the control unit generates the display data for displaying the maintenance target parts and the maintenance timing for each of the constituent devices as the maintenance information based on the estimation data. As a result, it becomes possible for the administrator of the blood purification device to easily check the maintenance information.

[0165] In the sixth embodiment of the present disclosure, in the third embodiment, the learned estimation model is generated by machine learning unified data of maintenance in the blood purification device, and estimates to perform maintenance performed in a predetermined period at the same timing. As a result, it is possible to reduce the number of maintenance times and increase the usage period of the blood purification device.

[0166] In the seventh embodiment of the present disclosure, in the third embodiment, the learned estimation model is generated by machine learning blood purification performance data of a plurality of the patients, and estimates including selection data of the next patient in the estimation data. As a result, it is possible to reduce the number of maintenance times and make the usage period of the blood purification device longer. In addition, since the administrator of the blood purification device can check data of a suitable next patient, it is possible to easily determine the assignment of patients to the blood purification device and the like.

[0167] In the eighth embodiment of the present disclosure, in the sixth or seventh embodiment, the learned estimation model estimates the maintenance timing of the blood purification device in units of a period, and estimates conditions for delaying the maintenance timing within the estimated period. As a result, it is possible to reduce the number of maintenance times and make the usage period of the blood purification device longer.

[0168] In the ninth embodiment of the present disclosure, in any of the first to eighth embodiments, the operation data acquisition unit includes a calculation unit that generates the operation data from the state parameters. Thereby, it becomes possible to easily acquire the operation data necessary for acquiring the estimation data.

[0169] In the tenth embodiment of the present disclosure, in the ninth embodiment, the calculation unit calculates the number of operations or the number of washings of the blood purification device from at least any one of the operation signal, temperature signal, pressure signal, or flow rate signal detected by the component device of the blood purification device or the sensor provided in the flow path of the blood purification device. Thereby, it becomes possible to obtain estimation data corresponding to the current operation state of the blood purification device.

[0170] In the eleventh embodiment of the present disclosure, in any of the first to tenth embodiments, the estimation unit inputs the past maintenance data of the blood purification device into the learned estimation model, and the estimation data is associated with the state change of the blood purification device from the time of past maintenance to the time of operation data acquisition. Thereby, the estimation accuracy by the learned estimation model can be improved, and it becomes possible to acquire more reliable estimation data.

[0171] In the twelfth embodiment of the present disclosure, in any of the first to eleventh embodiments, the control unit selects the order destination of the maintenance target component included in the estimation data from the order destination data and places an order for the maintenance target component. Thereby, the burden on the administrator of the blood purification device is reduced.

[0172] In the thirteenth embodiment of the present disclosure, in any of the first to twelfth embodiments, the operation data acquisition unit acquires the operation data of a plurality of the blood purification devices, and the control unit generates maintenance information for transmitting to a plurality of the blood purification devices or a portable terminal device of a maintenance performer. Thereby, it becomes possible for the administrator of the blood purification device to easily confirm the maintenance information.

[0173] In the 14th embodiment of the present disclosure, in the 13th embodiment, the learned estimation model is generated by performing machine learning while associating past usage status data, past failure data, past maintenance execution data, basic maintenance data, and blood purification performance data of a plurality of the patients of the blood purification device, and for each of the plurality of the blood purification devices, the next patient selection data is included in the estimation data for estimation. Thereby, since the administrator of the blood purification device can check the data of a suitable next patient, the allocation of patients to the blood purification device and the like can be easily determined.

[0174] A 15th embodiment of the present disclosure is an information processing method for estimating the maintenance timing of a blood purification device for purifying a patient's blood, including: a step of acquiring operation data generated from state parameters indicating the operation state of the blood purification device; a step of inputting the operation data into a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification device to obtain estimation data related to the next maintenance of the blood purification device; and a step of generating display data for displaying maintenance information of the blood purification device based on the estimation data. Thus, by simply inputting the operation data of the blood purification device into the learned estimation model, the maintenance timing of the maintenance target parts of the blood purification device is output, so that the information related to the maintenance can be easily estimated, and the frequency and timing of the maintenance can be optimized.

[0175] A 16th embodiment of the present disclosure is a program for estimating the maintenance timing of a blood purification device that purifies a patient's blood. The program acquires operation data generated from state parameters indicating the operation state of the blood purification device, inputs the operation data into a learned estimation model that has performed machine learning for estimating the maintenance timing of the blood purification device, acquires estimation data related to the next maintenance of the blood purification device, and generates display data for displaying maintenance information of the blood purification device based on the estimation data, and causes a computer to execute the process. In this way, by simply inputting the operation data of the blood purification device into the learned estimation model, the maintenance timing of the maintenance target parts of the blood purification device is output, so that information related to maintenance can be easily estimated, and the frequency and timing of the maintenance can be optimized.

Explanation of Signs

[0176] 1 Blood purification device 4 Display 4a Input section 4b Output section 5 Blood purifier 6 Information processing device 6a Processor 6b Memory 7 Internal piping section 8 Extracorporeal circulation section 9 Communication section 51 Operation data acquisition section 52 Estimation section 53 Control section 54 Storage section 55 Calculation section 56 Learned estimation model

Claims

1. An information processing apparatus for estimating the maintenance timing of a blood purification apparatus that purifies a patient's blood, comprising: an operation data acquisition unit that acquires operation data generated from state parameters indicating the operation state of the blood purification apparatus; an estimation unit that inputs the operation data into a learned estimation model obtained by performing machine learning for estimating the maintenance timing of the blood purification apparatus, and acquires estimation data related to the next maintenance of the blood purification apparatus; a control unit that generates display data for displaying maintenance information of the blood purification apparatus based on the estimation data, wherein the learned estimation model is generated by performing machine learning on past maintenance execution data of the blood purification apparatus.

2. The information processing apparatus according to claim 1, wherein the estimation data includes data of parts to be maintained of the blood purification apparatus.

3. The information processing apparatus according to claim 1, wherein the learned estimation model is generated by performing machine learning while associating past usage status data, past failure data, and basic maintenance data of the blood purification apparatus.

4. The information processing apparatus according to claim 1, wherein the state parameter includes at least one of an operation signal, a temperature signal, a pressure signal, or a flow rate signal detected by a component device constituting the blood purification apparatus or a sensor provided in a flow path of the blood purification apparatus.

5. The learned estimation model estimates parts to be maintained and maintenance timing for the parts to be maintained for each component device, and the control unit generates the display data for displaying the parts to be maintained and the maintenance timing for each component device as the maintenance information based on the estimation data.

6. The learned estimation model is generated by performing machine learning on unified maintenance data in the blood purification apparatus, and estimates that maintenance performed during a predetermined period is carried out at the same timing.

7. The learned estimation model is generated by performing machine learning on blood purification performance data of a plurality of the patients, and estimates including selection data of the next patient in the estimation data.

8. The information processing apparatus according to claim 6 or 7, wherein the learned estimation model estimates the maintenance timing of the blood purification apparatus in units of periods and estimates conditions for delaying the maintenance timing within the estimated periods.

9. The information processing apparatus according to claim 1, wherein the operation data acquisition unit includes a calculation unit that generates the operation data from the state parameters.

10. The information processing apparatus according to claim 9, wherein the calculation unit calculates the number of operations or the number of washings of the blood purification apparatus from at least any one of an operation signal, a temperature signal, a pressure signal, or a flow rate signal detected by a component device of the blood purification apparatus or a sensor provided in a flow path of the blood purification apparatus.

11. The estimation unit inputs past maintenance data of the blood purification apparatus into the learned estimation model. The information processing apparatus according to claim 1, wherein the estimation data is associated with a change in the state of the blood purification apparatus from the time of past maintenance implementation to the time of operation data acquisition.

12. The information processing apparatus according to claim 2, wherein the control unit selects an order destination of the maintenance target component included in the estimation data from order destination data and places an order for the maintenance target component.

13. The operation data acquisition unit acquires operation data of a plurality of the blood purification apparatuses. The information processing apparatus according to claim 1, wherein the control unit generates maintenance information for transmitting to a plurality of the blood purification apparatuses or a portable terminal device of a maintenance performer.

14. The learned estimation model is generated by performing machine learning by associating past usage status data, past failure data, past maintenance implementation data, basic maintenance data, and blood purification performance data of a plurality of the patients of the blood purification apparatus. The information processing apparatus according to claim 13, wherein the next patient selection data is included in the estimation data and estimated for each of the plurality of the blood purification apparatuses.

15. An information processing method for estimating the maintenance timing of a blood purification apparatus that purifies a patient's blood, the method including: a step of acquiring operation data generated from state parameters indicating an operation state of the blood purification apparatus; a step of inputting the operation data into a learned estimation model that has performed machine learning for estimating the maintenance timing of the blood purification apparatus, and acquiring estimation data related to the next maintenance of the blood purification apparatus. A step of generating display data for displaying maintenance information of the blood purification device based on the estimation data, The learned estimation model is generated by machine learning the past maintenance execution data of the blood purification device. An information processing method.

16. A program for estimating the maintenance timing of a blood purification device for purifying a patient's blood, Obtain operation data generated from state parameters indicating the operation state of the blood purification device, Input the operation data into a learned estimation model that has performed machine learning for estimating the maintenance timing of the blood purification device to obtain estimation data related to the next maintenance of the blood purification device, Cause a computer to execute a process of generating display data for displaying maintenance information of the blood purification device based on the estimation data, The learned estimation model is generated by machine learning the past maintenance execution data of the blood purification device. A program.

Images