Information processing device, information processing method, and program

Abstract

The ability to quantitatively identify odors related to a patient's medical condition from a patient's blood. [Solution] An information processing device for identifying the odor of a patient's blood, comprising: an acquisition unit that receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood and acquires odor data related to the predetermined odor from the detection signal; and an identification unit that identifies the odor related to the patient's condition or the status of blood purification treatment based on the odor data.

Description

Technical Field

[0001] The present disclosure relates to an information processing apparatus, an information processing method, and a program for identifying the odor of a patient's blood.

Background Art

[0002] Conventionally, it has been known that in the blood of a patient suffering from a disease, the concentration of a specific substance or component increases, and bad breath and body odor change. As an example, patients with renal failure have a high concentration of urea nitrogen (BUN) in their blood and have an ammonia odor from their bad breath. Also, as a means for detecting an odor, it is known that there is a sensor for detecting a specific odor (volatile substance) and measuring its concentration. As an example of the utilization of such a sensor, a fire alarm and an alcohol checker can be cited.

[0003] Patent Document 1 discloses a blood alcohol concentration measuring device capable of measuring the alcohol concentration from the blood or breath of a subject. In this blood alcohol concentration measuring device, the blood of a subject is collected at the accident scene, and the alcohol concentration in the collected blood is measured. This makes it possible to quickly measure the blood alcohol concentration at the accident scene, minimize the passage of time, and measure an accurate alcohol concentration.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, the concentration measuring devices described above can only measure and identify the blood alcohol concentration (alcohol odor) of the person being measured. In other words, these devices cannot identify odors related to pathological conditions such as whether or not the person being measured has a disease such as renal failure or diabetes. Furthermore, there was a problem in that odor-related data could not be continuously acquired and utilized with a single, one-off measurement.

[0006] This disclosure has been made in view of these issues, and its purpose is to provide an information processing device, an information processing method, and a program that can acquire odor-related data from the blood of a subject (patient) and quantitatively identify odors related to the patient's medical condition. [Means for solving the problem]

[0007] According to one aspect of the present disclosure, an information processing device for identifying the odor of a patient's blood is provided, comprising: an acquisition unit that receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood, and acquires odor data relating to the predetermined odor from the detection signal; and an identification unit that identifies the odor relating to the patient's condition or the status of blood purification treatment based on the odor data.

[0008] According to one aspect of this disclosure, "an information processing method for identifying the odor of a patient's blood using a computer is provided, comprising the steps of: receiving a detection signal from an odor sensor that detects a predetermined odor contained in the blood; acquiring odor data relating to the predetermined odor from the detection signal; and identifying an odor relating to the patient's condition or the status of blood purification treatment based on the odor data."

[0009] According to one aspect of this disclosure, a program is provided for identifying the odor of a patient's blood, which receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood, obtains odor data related to the predetermined odor from the detection signal, and causes a computer to perform processing to identify the odor related to the patient's medical condition or the status of blood purification treatment based on the odor data. [Effects of the Invention]

[0010] According to this disclosure, it is possible to provide an information processing device, an information processing method, and a program that can acquire odor-related data from the blood of a subject (patient) taken outside the body and quantitatively identify odors related to the patient's medical condition, etc.

[0011] The effects described above are merely illustrative for the sake of explanation, and the effects relating to this disclosure are not limited to those described above. In addition to the effects described above, any other effects described herein may be achieved. [Brief explanation of the drawing]

[0012] [Figure 1] This is a block diagram showing the electrical configuration of the information processing device according to the first embodiment. [Figure 2] This is a functional block diagram of the information processing system according to the first embodiment. [Figure 3] This is a schematic diagram showing an example of the use of the information processing system according to the first embodiment. [Figure 4] This is a data table stored in the information processing device according to the first embodiment. [Figure 5] This is a schematic diagram showing an example of the use of the information processing system according to the first embodiment. [Figure 6] This is a functional block diagram of the information processing system according to the second embodiment. [Figure 7] This is a flowchart showing the flow of setting up the trained identification model in the information processing device according to the second embodiment. [Figure 8]This is a schematic diagram showing information processing by a trained identification model set up in the information processing device according to the second embodiment. [Modes for carrying out the invention]

[0013] The information processing apparatus and information processing system including the same described herein will be described in detail below with reference to the drawings. Note that this disclosure is not limited to the content described below, and can be modified and implemented as such without altering its essence. Furthermore, the drawings used in each embodiment are schematic representations of the information processing apparatus, its components, and information processing system including them, and have been partially emphasized, enlarged, reduced, or omitted to enhance understanding, and may not accurately represent the scale or shape of each component. Additionally, some numerical values ​​used in each embodiment are examples only and can be changed as needed. Common components in the drawings are denoted by the same reference numerals.

[0014] <First Embodiment> (Configuration and processing of information processing equipment) The configuration of the information processing device and information processing system according to this disclosure, as well as the flow of information processing, will be explained with reference to Figures 1 to 4. Figure 1 is a block diagram showing the electrical configuration of the information processing device according to this embodiment. Figure 2 is a functional block diagram of the information processing system according to this embodiment, which also shows the flow of information and signals between each component. Figure 3 is a schematic diagram showing an example of use of the information processing system according to this embodiment. Figure 4 is a data table stored in the information processing device according to this embodiment.

[0015] As the information processing apparatus 1, typically, a wireless communication-capable apparatus typified by a large computer can be mentioned, but it is not of course limited to only such an apparatus. For example, as such an apparatus, any apparatus capable of executing the program according to the present disclosure, such as a feature phone, a portable information terminal, a PDA, a smartphone, a desktop personal computer, a laptop personal computer, a portable game machine, a stationary game machine, etc., can be preferably applied. Further, the information processing apparatus 1 does not necessarily need to include what is illustrated in FIG. 1 in a single housing, and it is also possible to distribute each component and process of the information processing apparatus 1 to a plurality of server apparatuses or cloud server apparatuses.

[0016] As shown in FIG. 1, the information processing apparatus 1 has an output interface 11, a processor 12, a memory 13, a communication interface 14 including a communication processing circuit 14a and an antenna 14b, and an input interface 15. And these respective components are electrically connected to each other via control lines and data lines. Note that the information processing apparatus 1 does not necessarily need to include all of the components shown in FIG. 1, and it is possible to take a configuration in which some are omitted, or it is also possible to add other components.

[0017] With the above configuration, in the information processing apparatus 1, transmission and reception of various signals, data, and information are possible. Note that the data is basically assumed to be composed of numerical values, symbols, characters, etc. obtained by processing signals and the like. Further, the information is basically assumed to be what is obtained by collecting or processing the data, and for example, it is assumed that the receiving side can use it as a material for subsequent consideration or can utilize it on the receiving side. However, regarding the data and information, depending on the content and the context before and after, there may be cases where they are used not in accordance with the above assumptions.

[0018] The output interface 11 has a function of outputting various displays output by executing the program according to the present disclosure in response to an instruction from the processor 12 to devices such as a display and a printer. Note that the display is composed of, for example, a liquid crystal display, an organic EL display, or an electronic paper.

[0019] The processor 12 is composed of a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), and has a function of controlling other connected components based on various programs stored in the memory 13. Specifically, the processor 12 reads out and executes a program for executing the information processing according to the present disclosure and a program for executing an OS from the memory 13. In the present embodiment, the processor 12 particularly executes the processing related to the information shown in FIGS. 2 and 4. Note that the processor 12 may be composed of a single CPU or GPU, or may be composed by combining a plurality of CPUs and GPUs.

[0020] In the present embodiment, the processor 12 receives a predetermined odor detection signal from the odor sensor 20 described later, and executes a process for identifying an odor related to the situation of a blood purification treatment (dialysis treatment) started by the operation of the blood purification unit and the patient's medical condition. Here, the situation of the blood purification treatment is, for example, a blood recirculation rate (VA recirculation rate), a dialysis volume (dialysis efficiency), and its change over time. Further, the odor related to the patient's medical condition is, for example, an odor related to the blood of a patient with renal failure or an odor related to the blood of a diabetic patient. These identification processes will be described later.

[0021] The memory 13 is composed of a ROM, a RAM, a non-volatile memory, an HDD, etc., and functions as a storage unit 22. The ROM stores instruction commands for executing the application and the OS according to the present disclosure as programs. The RAM is used for writing and reading data while the program stored in the ROM is being processed by the processor 12. The non-volatile memory is a memory 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.

[0022] Memory 13 stores a program for executing processing related to the information shown in Figures 2 and 4. More specifically, this program is a program for identifying the odor of a patient's blood, which receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood, obtains odor information related to the predetermined odor from the detection signal, and causes the information processing device 1 (i.e., processor 12), which is a computer, to execute processing for identifying the odor related to the patient's condition or the status of blood purification treatment based on the odor information. In this embodiment, memory 13 also stores various types of received information as shown in Figures 2 and 4.

[0023] Memory 13 stores a program for identifying odors related to the patient's condition or the status of blood purification therapy. In particular, memory 13 stores the type and numerical value of an odor set in accordance with a predetermined odor. Here, the type of odor refers to the type of odor, such as ammonia or acetone, and the numerical value of the odor is a numerical representation of the component concentration or odor index intensity of that odor. Memory 13 then associates these odor types, numerical values, and programs and stores them as odor information.

[0024] The communication interface 14 functions as a communication unit that sends and receives information with other terminal devices via the communication processing circuit 14a and antenna 14b. The communication interface 14 performs processing to send and receive information from other terminal devices as the processing of programs and various other information used in the information processing system 10 progresses.

[0025] The communication processing circuit 14a processes data based on a wideband wireless communication method such as LTE, but it can also process data based on a narrowband wireless communication method such as Wi-Fi (e.g., IEEE 802.11) or Bluetooth®, or a contactless wireless communication method. In addition, wired communication can be used instead of or in addition to wireless communication.

[0026] The input interface 15 has the function of receiving instruction inputs related to the execution of the program according to this embodiment, and operation inputs for registering various information. The input interface 15 may consist of, for example, a mouse, hard keys, or a touch panel, and if a touch panel is used, known methods such as resistive touch, capacitive coupling, and ultrasonic surface acoustic wave methods can be used.

[0027] As shown in Figure 2, the information processing system 10 consists of an information processing device 1, which is a device for identifying the odor of the patient's (subject's) blood, and an odor sensor 20 for detecting the odor of the patient's blood.

[0028] Furthermore, as shown in Figure 3, the odor sensor 20 is installed in the extracorporeal circulation section 5 of the blood purification unit (blood purification device) for circulating the patient's blood outside the body. The blood purification unit is a unit capable of taking the patient's blood outside the body (blood withdrawal), removing unnecessary or toxic substances or water from the blood in a blood purifier (blood purification), and returning the purified blood to the patient (blood return).

[0029] In this embodiment, the application of the information processing system 10 is described as being to a blood purification unit, but it is not limited to this. For example, devices for acute blood purification, peritoneal dialysis machines, ultrafiltration machines, or hemofiltration machines can also be applied to the information processing system 10. Furthermore, the specific configuration of the blood purification unit is not a feature of the information processing system described herein, so its description is omitted.

[0030] As shown in Figure 3, the arterial blood circuit L1, located in the extracorporeal circulation section 5 of the blood purification unit, is connected to a pump P1 and a chamber 3. The arterial blood circuit L1 is a circuit that introduces blood drawn from patient H into a blood purifier (not shown). A venous blood circuit (not shown) is connected to the blood purifier, and the blood purified by the blood purifier is returned to patient H. The arterial blood circuit L1 and the venous blood circuit constitute the blood circuit through which the patient's blood flows.

[0031] One end of the arterial blood circuit L1 is connected to a puncture needle (not shown) that is inserted into the arm of patient H, and the other end is connected to a blood purifier. In the arterial blood circuit L1, the connections are in the order of chamber 3 and pump P1 from the puncture needle side. When pump P1 rotates forward (clockwise), blood is drawn from patient H, and patient H's blood passes through the arterial blood circuit L1 to reach the blood purifier.

[0032] The type of pump P1 is not particularly limited, as long as it can introduce patient H's blood into the blood purifier. In this embodiment, a peristaltic pump is assumed, but other pumps such as diaphragm pumps may also be used.

[0033] An odor sensor 20 is connected to the chamber 3 on the arterial blood circuit L1. In this embodiment, the odor sensor 20 constitutes the extracorporeal circulation unit 5 and is provided as a consumable part of the blood purification unit, rather than being a component of the information processing device 1. The odor sensor 20 is also provided to detect a predetermined odor contained in the blood temporarily stored in the chamber 3. The odor sensor 20 may be of the type of semiconductor or quartz crystal oscillator. In addition, an analyzer such as a gas chromatograph that can quantify the concentration of chemical substances constituting the odor may be used as the odor sensor 20. The odor sensor 20 transmits the predetermined odor detected from the blood stored in the chamber 3 as a detection signal to the acquisition unit 21 of the information processing device 1.

[0034] As shown in Figure 2, the information processing device 1 includes an acquisition unit 21, a storage unit 22, an identification unit 23, an input unit 41, and an output unit 42. The acquisition unit 21, the storage unit 22, and the identification unit 23 are realized by the processor 12 and the memory 13 themselves functioning, or by the processor 12 reading and executing a program stored in the memory 13.

[0035] First, the acquisition unit 21 receives setting information from the input unit 41. Setting information is information used to construct a data table (see Figure 4) that is referenced when identifying the odor related to the patient's condition or the status of blood purification treatment. As an example, the data table stores "Type," "Value," "Blood Purification Treatment," and "Result" as setting information (odor information).

[0036] Here, "type" refers to information related to the type of odor, such as ammonia odor or acetone odor. "Value" refers to information that quantifies the concentration of odor components or the intensity of the odor index. In the data table in Figure 4, for example, "ammonia odor" is stored as the type for identifying the odor as being related to the blood of a patient with renal failure, and the value for this type is a predetermined value, "A1". This means that if the value for ammonia odor is A1 or higher, it is identified as the odor related to the blood of a patient with renal failure. Furthermore, "blood purification therapy" in the same data table refers to information related to the necessity of blood purification therapy in order to identify the odor related to the patient's condition or the status of blood purification therapy. For example, in order to identify "low dialysis efficiency" or "high blood recirculation rate", the patient must be undergoing blood purification therapy, so this is stored as "○" in the data table. On the other hand, in order to identify it as "the smell of blood from a patient with renal failure," it is not necessarily required that the patient be undergoing blood purification therapy. Identification is possible even with small amounts of blood collected for blood tests, etc., and therefore it is recorded as "—" in the data table.

[0037] Next, the acquisition unit 21 receives an odor detection signal from the odor sensor 20. Specifically, the acquisition unit 21 receives the odor of the patient's blood detected by the odor sensor 20 as a detection signal. Subsequently, the acquisition unit 21 acquires odor data from the detection signal and transmits the odor data to the identification unit 23. The odor data includes a predetermined type and numerical value of the odor, but may also include the date of detection and the detection environment. The odor data is used to identify the odor related to the patient's condition or the status of blood purification treatment by comparing it with the odor information in the data table in Figure 4.

[0038] Next, the identification unit 23 compares the odor data received from the acquisition unit 21 with the odor information read from the data table of the storage unit 22 to identify the odor related to the patient's condition or the status of blood purification treatment. Specifically, the identification unit 23 refers to the data table and determines whether the "type" included in the data table matches the type of odor included in the odor data, and also determines whether the numerical value of the odor included in the odor data is greater than or equal to the "numerical value" included in the data table. For example, if the type of odor related to the acquired odor data is acetone odor and its numerical value is A2 or higher, the identification unit 23 acquires "odor of blood of a diabetic patient" as the identification result.

[0039] The identification unit 23 then includes the identification result in the display information and transmits it to the output unit 42. As a result, the identification result is displayed on the output unit 42, and the operator of the information processing device 1 (a medical professional such as a doctor or nurse, or patient H) can check the odor related to the patient's condition or the status of blood purification treatment.

[0040] In addition to the configuration described above, the information processing device 1 may also have a communication unit that enables the transmission and reception of information with external devices such as terminal devices or server devices. In this case, the information processing device 1 may acquire various types of information related to the patient's treatment (patient information) from the external device via the communication unit.

[0041] In this embodiment, the specified odor may be, for example, an ammonia odor, acetone odor, indole odor, hydrogen sulfide odor contained in the patient's blood, an acetic acid odor contained in the dialysis fluid, or a unique odor of a blood purifier or blood circuit that does not occur in the body. The odor sensor 20 is capable of detecting the volatile substances that constitute these odors. The odor sensor 20 may also quantify the component concentration or odor index intensity of the detected odor.

[0042] Here, we will explain an example of using the ammonia odor in the blood to identify odors related to a patient's condition. First, as a premise, patient H has an elevated concentration of uremic toxins in his blood due to decreased renal function. In other words, patient H's blood has a higher ammonia odor level compared to a healthy person. Therefore, by detecting and utilizing the ammonia odor in the blood flowing through the arterial blood circuit L1, it is possible to identify the odor of the blood of a patient with decreased renal function (renal failure).

[0043] Furthermore, uremic toxins are removed from patient H's blood through blood purification therapy (dialysis). As a result, the concentration of uremic toxins in patient H's blood decreases over time as blood purification therapy progresses, and the ammonia odor level in the blood also decreases. In this way, the decrease in uremic toxin concentration as blood purification therapy progresses makes it possible to detect the change in ammonia odor level in the blood flowing through the arterial blood circuit L1 over time and use it as time-series data, which also allows for the identification of dialysis volume (dialysis efficiency).

[0044] As described above, in this example of use, the identification unit 23 can identify the odor of blood from patients with impaired renal function (renal failure) and the amount of dialysis (dialysis efficiency) in blood purification therapy (dialysis therapy) from the numerical value of ammonia odor in the blood flowing through the arterial blood circuit L1. In particular, when identifying the amount of dialysis, the identification result is reported to the operator of the information processing device 1, allowing the operator to confirm that a problem has occurred in the treatment and take early action.

[0045] In this embodiment, the identification unit 23 can identify odors related to the patient's condition not only from the numerical value of ammonia odor in the patient H's blood, but also from numerical values ​​related to other odors. For example, as shown in Figure 4, a high value of acetone odor in the blood indicates the odor of a diabetic patient's blood, a high value of indole odor indicates the odor of a constipated patient's blood, and a high value of hydrogen sulfide odor indicates the odor of a gastritis patient. By storing these values ​​as a data table in the storage unit 22, it can be used to identify odors related to the patient's condition.

[0046] Next, we will explain an example of using the acetic acid odor in the blood to identify the status of blood purification therapy. First, as a premise, the dialysate, which is the drug solution supplied to the blood purifier in blood purification therapy (dialysis therapy), generally contains an acetic acid odor. Therefore, during dialysis therapy, when the dialysate and blood come into contact via the blood purifier, the acetic acid odor contained in the dialysate is transferred to the blood, and blood with an acetic acid odor flows through the venous blood circuit. On the other hand, the blood flowing through the arterial blood circuit L1 is drawn from patient H, so it usually does not contain an acetic acid odor. However, when blood recirculation occurs, blood with an acetic acid odor flows through the arterial blood circuit L1. In other words, the value of acetic acid odor in the blood becomes high. Therefore, by detecting the change in the value of acetic acid odor in the blood flowing through the arterial blood circuit L1 over time and using it as time-series data, the blood recirculation rate can also be identified.

[0047] As described above, in this example of use, the identification unit 23 can identify the blood recirculation rate in blood purification therapy (dialysis therapy) from the numerical value of the acetic acid odor in the blood flowing through the arterial blood circuit L1. In this case, the identification result is notified to the operator of the information processing device 1, allowing the operator to confirm that a problem has occurred in the treatment and take early action.

[0048] Next, we will explain an example of using the dialysis-specific odor in the blood to identify the blood recirculation rate. First, as a premise, the blood purifier or blood circuit has a specific odor (i.e., a dialysis-specific odor). Therefore, during dialysis treatment, as blood passes through the blood purifier and blood circuit, the specific odor is gradually imparted to the blood, and this specific odor becomes stronger in the blood flowing through the venous blood circuit. On the other hand, the blood flowing through the arterial blood circuit L1 has a shorter passage distance through the blood circuit and has not passed through the blood purifier, so this specific odor is almost nonexistent or very weak. However, when blood recirculation occurs, blood with this specific odor flows through the arterial blood circuit L1. In other words, the value of this specific odor in the blood increases. Therefore, by detecting the change in the value of the dialysis-specific odor in the blood flowing through the arterial blood circuit L1 over time and using it as time-series data, the blood recirculation rate can also be identified.

[0049] As described above, in this example of use, the identification unit 23 can identify the blood recirculation rate in blood purification therapy (dialysis therapy) from the numerical value of the odor characteristic of dialysis therapy in the blood flowing through the arterial blood circuit L1. In this case, the identification result is notified to the operator of the information processing device 1, allowing the operator to confirm that a problem has occurred in the treatment and take early action.

[0050] (Effects of the first embodiment) In this embodiment, a detection signal is received from an odor sensor that detects a predetermined odor contained in the blood flowing through the blood circuit, and the odor related to the patient's condition or the status of blood purification therapy is identified from the detection signal. This allows for quantitative identification of the odor of the blood, leading to the early detection of diseases the patient has and early response to malfunctions in blood purification therapy.

[0051] (Modification of the first embodiment) In this embodiment, the odor sensor 20 for detecting the odor of blood was connected to the chamber 3 of the arterial blood circuit L1, but the location of the odor sensor 20 is not limited to this location. For example, the odor sensor 20 may be installed directly in the arterial blood circuit L1 instead of the chamber 3.

[0052] In this embodiment, the odor sensor 20 was provided as a consumable item in the arterial blood circuit L1, but the odor sensor 20 may also be a component of the information processing device 1. That is, the odor sensor 20 may be installed inside or on the surface of the information processing device 1.

[0053] In this embodiment, during treatment with the blood purification unit, odor data of the blood is acquired while the patient's blood is removed from the body, and the odor related to the patient's condition and the status of the blood purification treatment are identified. However, as shown in Figure 5, it is also possible to acquire odor data from the blood after it has been collected by the blood collector 51 and identify the odor related to the patient H's condition. In Figure 5, blood is collected from patient H using the blood collector 51 and injected into the blood collection tube 52. An odor sensor 20 is connected to the blood collection tube 52, and the acquisition unit 21 of the information processing device 1 receives the detection signal from the odor sensor 20. The subsequent processing is the same as described above, so its explanation is omitted. In this case, the odor of the blood can be quantitatively identified from the blood collected by blood tests, and only a small amount of blood needs to be removed from the body, thus reducing the burden on patient H.

[0054] In this embodiment, the identification unit 23 identified one of the following: the odor of blood from a patient with renal failure, the odor of blood from a patient with diabetes, the odor of blood from a patient with liver dysfunction, the odor of blood from a patient with constipation, the odor of blood from a patient with gastritis, a decrease in dialysis efficiency, or an increase in the blood recirculation rate. However, it may identify multiple odors simultaneously. For example, in Figure 4, if odor data is obtained in which the ammonia odor value in the blood is A1 or higher and the acetone odor value is A2 or higher, the identification unit 23 may identify it as "the odor of blood from a patient with renal failure and diabetes."

[0055] In this embodiment, the identification unit 23 definitively identified the odor related to the patient's condition and the status of blood purification treatment, but it may also identify these with probability and confidence. For example, when identifying the odor related to the blood of a patient with renal failure, the "probability that it is the odor related to the blood of a patient with renal failure" may be stored in the data table for each of several numerical values ​​related to ammonia odor. Specifically, for example, if the identification unit 23 obtains odor data where the ammonia odor value is A1 or higher, it may identify that "the probability that it is the odor related to the blood of a patient with renal failure is 15%", and if numerical values ​​higher than A1 are stored in the data table in stages, it may identify that "the probability that it is the odor related to the blood of a patient with renal failure is 20%", "the probability that it is the odor related to the blood of a patient with renal failure is 30%", and so on.

[0056] <Second Embodiment> In the first embodiment, the patient's condition, odor, or the status of blood purification treatment was identified by using a data table. However, artificial intelligence (AI) may be used to identify these. This case will be described as the second embodiment with reference to Figures 6 to 8. Here, Figure 6 is a functional block diagram of the information processing system according to this embodiment. Figure 7 is a flowchart showing the flow of setting the trained identification model to be set in the information processing device according to this embodiment. Figure 8 is a schematic diagram showing information processing by the trained identification model to be set in the information processing device according to this embodiment. Note that only the parts that differ from the first embodiment will be described in detail, and the same content will be omitted from the explanation, and the same reference numerals will be used in the drawings.

[0057] As shown in Figure 6, in the information processing system 110 according to this embodiment, the identification unit 123 of the information processing device 101 has a trained identification model 124. The identification unit 123 inputs odor data received from the acquisition unit 21 into the trained identification model 124 in order to identify the odor of blood using the trained identification model 124. The trained identification model 124 uses artificial intelligence to identify the odor of blood from the input data and outputs the identification result, which is the odor related to the patient's condition or the status of blood purification treatment. When the identification unit 123 acquires the output data (identification data) of the trained identification model 124, it transmits the data to the output unit 42 as display information.

[0058] During the generation and implementation of the trained identification model 124, the information processing device 101 undergoes initial setup. Here, initial setup refers to the preparatory processing necessary to identify the odor of blood using AI. In other words, the initial setup generates the trained identification model 124, and the trained identification model 124 is implemented in the information processing device 101.

[0059] Specifically, as shown in Figure 7, the processor 12 of the information processing device 101 acquires training data for generating a trained identification model 124 (S101). Here, as an example of training data, as shown in Figure 8, odor data related to the blood of healthy individuals, odor data related to the blood of patients with renal failure, odor data related to the blood of diabetic patients, odor data related to the blood of cancer patients, and odor data related to the blood undergoing blood purification therapy may be used. Furthermore, this data may be input by the administrator of the information processing device 101 via the input unit 41, or received via the communication unit.

[0060] Odor data for the blood of healthy individuals consists of odor data from blood samples taken from multiple healthy individuals who are not suffering from any disease. Odor data for the blood of patients with renal failure consists of odor data from blood samples taken from multiple patients with renal failure. Odor data for the blood of patients with diabetes consists of odor data from blood samples taken from multiple patients with diabetes. Odor data for the blood of cancer patients consists of odor data from blood samples taken from multiple cancer patients. Odor data for blood undergoing blood purification therapy consists of odor data from blood samples taken from multiple patients undergoing blood purification therapy, and may include information related to dialysis volume and blood circulation rate. Furthermore, this odor data may include information related to the distribution of odors, and may also include information that quantifies the concentration of odor components or the intensity of the odor index.

[0061] As described above, in this embodiment, odor data from the blood of multiple healthy individuals, odor data from the blood of multiple patients with renal failure, odor data from the blood of multiple patients with diabetes, odor data from the blood of multiple cancer patients, and odor data from blood undergoing blood purification therapy are associated and used in order to generate the trained discrimination model 124. While it is preferable that all of this data be associated, some data may be excluded. On the other hand, additional data may be added, and the associations between the data may be appropriately modified according to the machine learning described later.

[0062] Next, the processor 12 of the information processing device 101 performs annotation processing on the acquired training data (S102). Specifically, the annotation processing involves adding annotations to the acquired training data to generate training data. For example, the processor 12 tags standard odor data and generates correct answer data to which that odor data is associated.

[0063] Next, the processor 12 of the information processing device 101 performs machine learning using the acquired training data and annotated training data (S103). As an example, this machine learning is performed by providing training data and training data to a neural network composed of combinations of neurons, and repeatedly adjusting the parameters of each neuron so that the output of the neural network matches the correct data from the training data. Note that the above machine learning is merely an example, and machine learning using scoring may also be performed.

[0064] Next, the processor 12 of the information processing device 101 evaluates the trained identification model 124 generated by machine learning (S104). Here, the processor 12 uses evaluation data different from the training data used for machine learning to evaluate the trained identification model 124. For example, as shown in Figure 8, the processor 12 inputs data related to the type and numerical value of the odor as evaluation data into the trained identification model 124. Then, the processor 12 evaluates whether the blood odor or blood purification treatment status related to the patient's condition, which is output as identification data from the trained identification model 124, is a correct result. As a specific evaluation method, the processor 12 determines whether the blood odor or blood purification treatment status related to the actual patient's condition corresponding to the evaluation data matches the blood odor or blood purification treatment status related to the patient's condition corresponding to the identification data. If they do not match, the process is restarted from acquiring the training data, and machine learning is performed again.

[0065] If the above conditions are met, the processor 12 of the information processing device 101 implements the generated trained identification model 124 (S105). Specifically, the processor 12 stores the generated trained identification model 124 in the memory 13. As a result, the identification unit 123 in the information processing device 101, which has the trained identification model 124, becomes functional.

[0066] (Effects of the second embodiment) In this embodiment as well, a detection signal is received from an odor sensor that detects a predetermined odor contained in the blood flowing through the blood circuit, and the odor related to the patient's condition and the status of blood purification treatment are identified from the detection signal. In particular, in this embodiment, when data related to the type and numerical value of the odor is input to a trained identification model 124, which is generated by machine learning by associating odor data related to the blood of healthy individuals, odor data related to the blood of patients with renal failure, odor data related to the blood of diabetic patients, odor data related to the blood of cancer patients, and odor data related to the blood undergoing blood purification treatment, the trained identification model 124 outputs the odor related to the patient's condition or the status of blood purification treatment. Since the trained identification model 124 is generated by machine learning by associating the various data mentioned above, it is possible to output identification results with higher reliability. Furthermore, according to the above machine learning, it is possible to identify odors in which odor substances such as cancer odor have not been identified.

[0067] (Modified version of the second embodiment) In this embodiment, during treatment with the blood purification unit, odor data of the blood is acquired while the blood is removed from the body, and the odor related to the patient's condition and the status of the blood purification treatment are identified. However, similar to the first embodiment, it is also possible to acquire odor data from the blood after it has been collected with a blood collection device and identify the odor related to the patient's condition. In this case, the odor of the blood can be quantitatively identified from the blood collected by blood tests, etc., and only a small amount of blood needs to be removed from the body, thus reducing the burden on the patient.

[0068] In this embodiment as well, the location of the odor sensor 20 is not particularly limited and may be connected to the chamber of the arterial blood circuit, or it may be installed directly in the arterial blood circuit instead of in the chamber.

[0069] In this embodiment as well, the odor sensor 20 may be provided in the arterial blood circuit as a consumable item, or the odor sensor 20 may be a component of the information processing device 101. That is, the odor sensor 20 may be installed inside or on the surface of the information processing device 101.

[0070] In this embodiment as well, the identification unit 123 may identify multiple odors simultaneously. For example, if odor data such as high values ​​for both ammonia odor and acetone odor in the blood is input to the trained identification model 124, the trained identification model 124 may output identification data indicating that it is "odor related to the blood of a patient with renal failure and diabetes."

[0071] In this embodiment as well, the identification unit 123 may identify odors related to the patient's condition and the status of blood purification treatment with probability and confidence. For example, when identifying the odor related to the blood of a patient with renal failure, the trained identification model 124 may output the "probability that it is the odor related to the blood of a patient with renal failure" as identification data. Specifically, if odor data indicating a high ammonia odor value is input to the trained identification model 124, the trained identification model 124 may output identification data indicating that "the probability that it is the odor related to the blood of a patient with renal failure is 15%."

[0072] <Embodiments of this disclosure> A first embodiment of the present disclosure is an information processing device for identifying the odor of a patient's blood, comprising: an acquisition unit that receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood and acquires odor data related to the predetermined odor from the detection signal; and an identification unit that identifies the odor related to the patient's condition or the status of blood purification treatment based on the odor data.

[0073] This allows for the acquisition of odor-related data from a patient's blood extracted from the body, enabling quantitative identification of odors related to the patient's condition and the status of blood purification therapy. Furthermore, identifying the odor of the blood can lead to the early detection of diseases the patient may have and early intervention for problems with blood purification therapy.

[0074] A second embodiment of this disclosure is, in the first embodiment, that the identification unit identifies the dialysis volume or blood recirculation rate based on the change in the numerical value of a predetermined odor over time during the blood purification treatment. This allows odor data related to the status of the blood purification treatment to be acquired and used as time-series data. Furthermore, the treatment status can be notified to the operator of the blood purification unit, leading to early response to treatment malfunctions.

[0075] A third embodiment of this disclosure is, in the second embodiment, having a database that associates the predetermined odor type and numerical value with the odor related to the patient's condition and the status of the blood purification treatment, wherein the identification unit refers to the database to identify the odor related to the patient's condition or the status of the blood purification treatment from the acquired odor data. This makes it possible to identify the odor related to the patient's condition and the status of the blood purification treatment, leading to the early detection of diseases the patient has and early response to malfunctions in the blood purification treatment.

[0076] A fourth embodiment of this disclosure is that, in the third embodiment, the odor data includes the predetermined odor type and numerical value. This makes it possible to identify odors related to the patient's condition and the status of blood purification therapy, leading to early detection of diseases the patient has and early response to malfunctions in blood purification therapy.

[0077] A fifth embodiment of this disclosure is, in the first embodiment, that the identification unit inputs the odor data into a trained identification model that has been machine-learned to identify the odor related to the patient's condition or the status of the blood purification treatment, thereby obtaining identification data related to the odor related to the patient's condition or the status of the blood purification treatment. This improves the accuracy of identification related to the odor related to the patient's condition or the status of the blood purification treatment using the trained identification model, enabling more reliable identification.

[0078] A sixth embodiment of this disclosure is that, in the fifth embodiment, the trained identification model is generated by machine learning by associating odor data related to the blood of healthy individuals, odor data related to the blood of patients with renal failure, odor data related to the blood of patients with diabetes, odor data related to the blood of cancer patients, and odor data related to the blood undergoing blood purification therapy. This improves the accuracy of identifying odors related to the patient's condition or the status of the blood purification therapy, enabling more reliable identification.

[0079] A seventh embodiment of this disclosure is that any of the first to sixth embodiments includes the odor sensor. This makes the odor sensor part of the configuration of the information processing device, allowing for more precise control of the odor sensor and more accurate reception of detection signals from the odor sensor.

[0080] An eighth embodiment of this disclosure is that, in the first embodiment, the patient's blood is collected by a blood collection device or by a blood purification device during blood purification therapy. This makes it possible to identify the odor related to the patient's condition and the status of the blood purification therapy, regardless of whether the blood is collected during blood purification therapy, during blood donation processing, or for blood tests, etc.

[0081] A ninth embodiment of this disclosure is an information processing method for identifying the odor of a patient's blood using a computer, comprising the steps of: receiving a detection signal from an odor sensor that detects a predetermined odor contained in the blood; acquiring odor data related to the predetermined odor from the detection signal; and identifying the odor related to the patient's condition or the status of blood purification treatment based on the odor data. This makes it possible to acquire odor-related data from a patient's blood taken outside the body and quantitatively identify the odor related to the patient's condition and the status of blood purification treatment. Furthermore, identifying the odor of the blood can lead to the early detection of diseases the patient has and to early response to malfunctions in blood purification treatment.

[0082] A tenth embodiment of this disclosure is a program for identifying the odor of a patient's blood, which receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood, obtains odor data related to the predetermined odor from the detection signal, and causes a computer to perform processing to identify the odor related to the patient's condition or the status of blood purification therapy based on the odor data. This makes it possible to obtain odor data from a patient's blood taken outside the body and quantitatively identify the odor related to the patient's condition and the status of blood purification therapy. Furthermore, by identifying the odor of the blood, it is possible to lead to the early detection of diseases the patient has and to early response to malfunctions in blood purification therapy. [Explanation of Symbols]

[0083] 1. Information Processing Device 10 Information Processing Systems 20 Odor Sensor 21 Acquisition Department 22 Memory section 23 Identification unit 41 Input section 42 Output section 124 Pre-trained Discrimination Models L1 arterial blood circuit P1 Pump H patient

Claims

1. An information processing device for identifying the odor of a patient's blood, An acquisition unit receives a detection signal from an odor sensor that detects a predetermined odor contained in the blood, and acquires odor data related to the predetermined odor from the detection signal. An information processing device having an identification unit that identifies the odor related to the patient's condition or the status of blood purification treatment based on the odor data.

2. The information processing apparatus according to claim 1, wherein the identification unit identifies the dialysis volume or blood recirculation rate based on the change in the numerical value of a predetermined odor over time as the blood purification treatment progresses.

3. The system has a database that associates the predetermined odor type and numerical value with the odor related to the patient's condition and the status of the blood purification treatment. The information processing apparatus according to claim 2, wherein the identification unit refers to the database and identifies the odor related to the patient's condition or the status of the blood purification treatment from the acquired odor data.

4. The information processing apparatus according to claim 3, wherein the odor data includes the predetermined type of odor and a numerical value.

5. The information processing apparatus according to claim 1, wherein the identification unit inputs the odor data into a trained identification model that has been trained using machine learning to identify the odor related to the patient's medical condition or the status of the blood purification treatment, and acquires identification data related to the odor related to the patient's medical condition or the status of the blood purification treatment.

6. The information processing apparatus according to claim 5, wherein the trained identification model is generated by machine learning by associating odor data related to the blood of healthy individuals, odor data related to the blood of patients with renal failure, odor data related to the blood of patients with diabetes, odor data related to the blood of cancer patients, and odor data related to the blood undergoing blood purification therapy.

7. The information processing apparatus according to any one of claims 1 to 6, comprising the odor sensor.

8. The information processing apparatus according to claim 1, wherein the patient's blood is collected by a blood collection device or by a blood purification device during blood purification therapy.

9. A method for processing information to identify the odor of a patient's blood using a computer, The process includes receiving a detection signal from an odor sensor that detects a predetermined odor contained in the blood, and obtaining odor data related to the predetermined odor from the detection signal, An information processing method comprising the step of identifying the odor related to the patient's medical condition or the status of blood purification treatment based on the odor data.

10. A program that identifies the odor of a patient's blood, A detection signal is received from an odor sensor that detects a predetermined odor contained in the blood, and odor data related to the predetermined odor is obtained from the detection signal. A program that causes a computer to perform processing to identify the odor related to the patient's medical condition or the status of blood purification treatment based on the odor data.

Images