Sensing system and sensing method

Abstract

To provide a sensing system, a learning apparatus, a sensing method, and a learning method, capable of deriving a flow rate from pressure.SOLUTION: A sensing system comprises: a sensor module that stores pressure data acquired by sensing pressure caused at a subject's predetermined part; and an information processor that acquires the pressure data stored by the sensor module and subject identification information, and uses a learning model created by performing machine learning of relation between information identifying pressure and information identifying a flow rate to derive flow rate data from the acquired pressure data for each subject identification information.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a sensing system, a learning device, a sensing method, and a learning method. [Background technology]

[0002] In a fluid, there is a strong correlation between pressure and the flow rate of the fluid, as shown by Bernoulli's theorem. However, there are strong constraints for the strong correlation between pressure and the flow rate of the fluid. The constraints are that the fluid is incompressible and that comparisons can only be made on the same streamline (Bernoulli surface). In particular, the fluid must be a vortex-free flow. Known methods of calculating flow rate from pressure using Bernoulli's theorem include the Venturi tube and Pitot tube. The Venturi tube stabilizes the state of the fluid by flowing it through a tube of known cross-sectional area for a sufficient inlet distance, and smoothly changes the cross-sectional area along the way. The Venturi tube calculates the flow velocity by measuring the pressure before and after the change in cross-sectional area, and derives the flow rate at that point. The Pitot tube calculates the flow velocity by measuring the pressure difference between the total pressure and static pressure of the flow. The Pitot tube can also derive the flow rate. Regarding the technology using the flow rate of a fluid, a technology is known that can easily detect a heartbeat signal that indicates the cardiac output of a living body without using ECG electrodes attached to the living body (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2018-202195 A Summary of the Invention [Problem to be solved by the invention]

[0004] To calculate the flow rate from the pressure, it is necessary that the conditions satisfy Bernoulli's theorem. The present invention has been made in consideration of the above problems, and has an object to provide a sensing system, a learning device, a sensing method, and a learning method that can derive a flow rate from a pressure. [Means for solving the problem]

[0005] (1) In view of the above-mentioned problems, a sensing system according to one aspect of the present invention is a sensing system that senses biological information using a sensor module, the sensing system including: a sensor module that stores pressure data acquired by sensing pressure occurring at a predetermined part of a subject; and information that acquires the pressure data and identification information of the subject stored by the sensor module and specifies the pressure. Pressure information and flow rate identification information Flow information and an information processing device that derives flow rate data for each identification information of the subject from the acquired pressure data using a learning model created by machine learning the relationship between the pressure data and the identification information of the subject. The sensor module is installed in the nasal cavity. The learning model is created by dividing the pressure information and the flow rate information sampled for each predetermined section into blocks for at least one breath section, and learning the relationship between the pressure information and the flow rate information for each block based on the blocking results, and has information overlap in which the pressure information and the flow rate information at a certain time are included in both a first block and a second block that are adjacent to each other in the time axis direction among the blocks. . (2) In view of the above-mentioned problems, a sensing system according to one aspect of the present invention is a sensing system that senses biological information using a sensor module, the sensing system including: a sensor module that stores pressure data acquired by sensing pressure occurring at a predetermined part of a subject; and information that acquires the pressure data and identification information of the subject stored by the sensor module and specifies the pressure. Pressure information and flow rate identification information Flow information and an information processing device that derives flow rate data for each subject's identification information from the acquired pressure data using a learning model created by machine learning the relationship between The pressure data includes static pressure only. . (3) In view of the above-mentioned problems, a sensing system according to one embodiment of the present invention includes a sensor module comprising a sensor for sensing pressure, a battery for supplying power, a signal processing unit for creating digital data based on the results of sensing by the sensor, a memory for storing the digital data created by the signal processing unit, and a wireless transceiver unit that receives a command to acquire data transmitted by an acquisition device, and transmits the digital data stored in the memory and the identification information of the subject to the acquisition device based on the received command. (4) In a sensing system according to one aspect of the present invention, the sensor Sa is , sensing the pressure in the left nasal cavity and / or the pressure in the right nasal cavity death, The information processing device uses the learning model to derive either or both of the flow rate in the left nasal cavity and the flow rate in the right nasal cavity from either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity. (5) In a sensing system according to an aspect of the present invention, the sensor module further includes at least one of a temperature sensor, a humidity sensor, an acceleration sensor, and a pulse wave sensor, and the learning model further includes, based on at least one of information specifying a temperature, information specifying a humidity, information specifying an acceleration, and information specifying a pulse wave, The pressure information and The flow rate information It was created by machine learning the relationship between (6) A sensing method according to one aspect of the present invention is a sensing method executed by a sensing system that senses biological information using a sensor module, the sensing method including the steps of: storing pressure data acquired by sensing pressure occurring at a predetermined part of a subject; acquiring the pressure data stored by the sensor module and identification information of the subject; and storing information for identifying the pressure. Pressure information and flow rate identification information Flow information and deriving flow rate data for each identification information of the subject from the acquired pressure data using a learning model created by machine learning the relationship between the pressure data and the identification information of the subject. The sensor module is installed in the nasal cavity. The learning model is created by dividing the pressure information and the flow rate information sampled for each predetermined section into blocks for at least one breath section, and learning the relationship between the pressure information and the flow rate information for each block based on the blocking results, and has information overlap in which the pressure information and the flow rate information at a certain time are included in both a first block and a second block that are adjacent to each other in the time axis direction among the blocks. . (7) A sensing method according to one aspect of the present invention is a sensing method executed by a sensing system that senses biological information using a sensor module, the sensing method including the steps of: storing pressure data acquired by sensing pressure occurring at a predetermined part of a subject; acquiring the pressure data stored by the sensor module and identification information of the subject; and storing information for identifying the pressure. Pressure information and flow rate identification information Flow information and deriving flow rate data from the acquired pressure data for each subject's identification information using a learning model created by machine learning the relationship between The sensor module is placed in the nasal cavity and the pressure includes static pressure only pressure data. Effect of the Invention

[0006] According to the present invention, it is possible to provide a sensing system, a learning device, a sensing method, and a learning method that can derive a flow rate from a pressure. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a diagram illustrating an example of a sensing system according to an embodiment of the present invention. [Diagram 2] 2 is a block diagram showing an example of a sensor module, a data acquisition device, and an information processing device included in the sensing system of the present embodiment. FIG. [Diagram 3] FIG. 2 is a diagram illustrating an example of a sensor module of the sensing system of the present embodiment. [Figure 4] FIG. 4 is a flow diagram showing a first example of the operation of the sensing system of the present embodiment. [Diagram 5] FIG. 11 is a flow diagram showing a second example of the operation of the sensing system of the present embodiment. [Figure 6] FIG. 1 is a diagram illustrating an example of a learning device according to an embodiment of the present invention. [Figure 7] FIG. 13 is a diagram showing example 1 of input data. [Figure 8] FIG. 11 is a diagram illustrating a second example of input data. [Figure 9]FIG. 11 is a diagram illustrating an example 3 of input data. [Figure 10] FIG. 4 is a diagram for explaining an example of processing of the learning device according to the present embodiment. [Figure 11] FIG. 4 is a diagram showing an example of output data of the information processing device according to the embodiment. [Figure 12] FIG. 4 is a diagram illustrating an example of the operation of the learning device according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Next, a sensing system, a learning device, a sensing method, and a learning method according to the present embodiment will be described with reference to the drawings. The embodiment described below is merely an example, and the embodiment to which the present invention is applied is not limited to the following embodiment. In addition, "based on XX" in this application means "based on at least XX," and includes cases where it is based on other elements in addition to XX. Furthermore, "based on XX" is not limited to cases where XX is used directly, but also includes cases where it is based on XX that has been calculated or processed. "XX" is any element (for example, any information).

[0009] (Embodiment) Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Components having the same or similar functions are denoted by the same reference numerals, and duplicate descriptions of the components may be omitted. (Sensing System) FIG. 1 is a diagram showing an example of a sensing system according to an embodiment of the present invention. The sensing system 100 of this embodiment monitors biological information using a sensor module installed at a predetermined part of a subject. The sensing system 100 includes a sensor module 1, a data acquisition device 2, and an information processing device 3. The sensor module 1 and the data acquisition device 2 are connected wirelessly. The data acquisition device 2 and the information processing device 3 communicate with each other via a network NW. The network NW includes, for example, the Internet, a Wide Area Network (WAN), a Local Area Network (LAN), a provider device, a wireless base station, and the like.

[0010] The sensor module 1 is placed at a predetermined site of the subject HS, and stores data acquired by sensing biometric information in association with information on the date and time when the data was acquired. An example of the predetermined site is the inside of the nasal cavity. Below, the explanation will continue for the case where the inside of the nasal cavity is used as an example of the predetermined site. An example of biometric information is information specifying intranasal pressure. Below, the explanation will continue for the case where information specifying intranasal pressure is used as an example of biometric information. The data acquisition device 2 creates a command indicating that sensor information is to be acquired, and transmits the created command to the sensor module 1. When the sensor module 1 receives a command indicating to acquire sensor information transmitted by the data acquisition device 2, it creates sensor information including the stored data, date and time information, and identification information of the subject HS (hereinafter referred to as "subject ID"), and transmits the created sensor information to the data acquisition device 2. The data acquisition device 2 receives the sensor information transmitted by the sensor module 1 in response to the transmitted command. The data acquisition device 2 transmits the received sensor information to the information processing device 3. The information processing device 3 acquires the data, date and time information, and subject ID included in the sensor information transmitted by the data acquisition device 2. The information processing device 3 analyzes the acquired data based on the date and time information. Specifically, the information processing device 3 uses a learning model to derive information specifying the nasal respiratory flow rate from information specifying the intranasal pressure.

[0011] The sensor module 1, the data acquisition device 2, and the information processing device 3 included in the sensing system 100 will be described in detail below. FIG. 2 is a block diagram showing an example of a sensor module, a data acquisition device, and an information processing device included in the sensing system of this embodiment. (Sensor module 1) The sensor module 1 includes a battery 11, a sensor 12, a sensor module section 13, a path section R, and a nose gripping section NG. Circuit blocks such as the battery 11, the sensor 12, and the sensor module section 13 are mounted on a thin circuit board or a flexible board. An example of the sensor module section 13 is realized by a chip including a signal processing section 14, a memory 15, and a wireless transmitting / receiving section 16. An example of the sensor module 1 is attached inside the nasal cavity by the nose gripping section NG.

[0012] FIG. 3 is a diagram illustrating an example of a sensor module of the sensing system of the present embodiment. 3, the battery 11 and the sensor module portion 13 included in the sensor module 1 are omitted. The sensor 12 is connected to a path portion R, and the path portion R is connected to the nose gripping portion NG. The path portion R has a left path portion RL and a right path portion RR. The sensor 12 has a left sensor 12L and a right sensor 12R. The nose gripping portion NG has a left nose gripping portion NGL and a right nose gripping portion NGR. The left nose gripping portion NGL is attached to the left nostril NCL, and the right nose gripping portion NGR is attached to the right nostril NCR, and the nose gripping portion NG can be attached to the nasal cavity by sandwiching the nasal septum between the left nose gripping portion NGL and the right nose gripping portion NGR. The left nose gripping portion NGL has a left cavity portion CL, and the right nose gripping portion NGR has a right cavity portion CR. The left cavity portion CL is connected to the left sensor 12L via the left path portion RL. The right cavity portion CR is connected to the right sensor 12R via the right path portion RR.

[0013] When the nose gripper NG is attached to the nasal cavity, a portion of the inhaled air flows from the left sensor 12L through the left path RL to the left cavity CL, and from the right sensor 12R through the right path RR to the right cavity CR. The left sensor 12L measures the pressure generated by the portion of the inhaled air flowing in the left nasal cavity. The right sensor 12R measures the pressure generated by the portion of the inhaled air flowing in the right nasal cavity. When the nose gripper NG is attached to the nasal cavity, a portion of the exhaled air flows from the left cavity portion CL through the left path portion RL to the left sensor 12L, and also flows from the right cavity portion CR through the right path portion RR to the right sensor 12R. The left sensor 12L measures the pressure generated by the portion of the exhaled air flowing in the left nasal cavity. The right sensor 12R measures the pressure generated by the portion of the exhaled air flowing in the right nasal cavity. Returning to FIG. 2, the explanation will be continued.

[0014] An example of the battery 11 is a primary battery or a secondary battery. The battery 11 supplies power to the sensor 12 and the sensor module unit 13. The sensors 12 (left sensor 12L, right sensor 12R) include pressure sensors, and acquire sensing data (pressure measurement results) by sensing. The sensors 12 output the acquired sensing data to the sensor module unit 13. Here, the sensing data (sensor data) is an output value of the sensor 12 mounted on the sensor module 1. For example, if the sensor module 1 is mounted with a pressure sensor, the pressure sensor outputs the pressure sensor value at each measurement interval.

[0015] The wireless transceiver 16 communicates with the data acquisition device 2. An example of a wireless communication method used for communication between the wireless transceiver 16 and the data acquisition device 2 is Bluetooth Low Energy (BLE). However, wireless communication between the wireless transceiver 16 and the data acquisition device 2 may be performed using a communication method other than BLE. Specifically, the wireless transceiver 16 receives the initial setting information transmitted by the data acquisition device 2. The initial setting information includes date and time information when the sensor module 1 starts sensing, information indicating the measurement time interval, and the subject ID. Here, the date and time information may include information indicating the year and month. Furthermore, the wireless transceiver 16 receives a command transmitted by the data acquisition device 2. The wireless transceiver 16 acquires sensor information output by the signal processing unit 14 in response to the received command, and transmits the acquired sensor information to the data acquisition device 2.

[0016] The signal processing unit 14 includes a microcomputer and an analog to digital converter (ADC). The signal processing unit 14 acquires the initial setting information received by the wireless transceiver unit 16, and acquires the date and time information for starting sensing, information indicating the measurement time interval, and the subject ID, which are included in the acquired initial setting information. The signal processing unit 14 stores the acquired date and time information for starting sensing, information indicating the measurement time interval, and the subject ID in the memory 15. The signal processing unit 14 also sets the acquired information indicating the measurement time interval in the sensor 12. The sensor 12 performs sensing at the measurement time interval based on information indicating the measurement time interval set by the signal processing unit 14. The signal processing unit 14 acquires the sensing data output by the sensor 12 and converts the acquired sensing data into digital data. The signal processing unit 14 associates the digital data obtained by converting the sensing data with a measurement index and stores the data in the memory 15. Here, the measurement index is identification information for the digital data. The measurement index and the sensor data are saved for each measurement interval. One example of the measurement index is a unique numerical value assigned in the order of measurement.

[0017] The signal processing unit 14 acquires a command received by the wireless transceiver unit 16 and interprets the acquired command. The signal processing unit 14 controls the memory 15 and the wireless transceiver unit 16 based on the result of interpreting the command. Specifically, when the command received by the wireless transceiver unit 16 is to acquire sensor information, the signal processing unit 14 creates sensor information addressed to the data acquisition device 2, including date and time information for starting sensing stored in the memory 15, information indicating the time interval for measurement, information associating the subject ID, the ID of the sensor module 1 (the ID of the sensor 12), the measurement index, and digital data. Here, the ID of the sensor module 1 is identification information of the sensor module 1. The signal processing unit 14 outputs the created sensor information to the wireless transceiver unit 16.

[0018] (Data acquisition device 2) The data acquisition device 2 is realized by a device such as a personal computer, a server, a smartphone, a tablet computer, an industrial computer, etc. The data acquisition device 2 includes, for example, a first communication unit 21-1, a second communication unit 21-2, a processing unit 22, and a storage unit 25. The first communication unit 21-1 is realized by a communication module. The first communication unit 21-1 communicates with an external communication device. The first communication unit 21-1 may communicate using a communication method such as BLE. The first communication unit 21-1 holds communication information required to communicate with the sensor module 1. The first communication unit 21-1 transmits initial setting information output by the processing unit 22 to the sensor module 1. The first communication unit 21-1 transmits a command output by the processing unit 22 to the sensor module 1. The first communication unit 21-1 receives sensor information transmitted by the sensor module 1 in response to the transmitted command. The second communication unit 21-2 is realized by a communication module. The second communication unit 21-2 communicates with an external communication device via the network NW. The second communication unit 21-2 may communicate using a communication method such as a wired LAN. The second communication unit 21-2 holds communication information required to communicate with the information processing device 3 via the network NW. The second communication unit 21-2 receives a sensor information acquisition request transmitted by the information processing device 3. The second communication unit 21-2 receives a sensor information request transmitted by the information processing device 3. The second communication unit 21-2 transmits a sensor information response output by the processing unit 22 to the information processing device 3.

[0019] The storage unit 25 is realized by a hard disk drive (HDD), a flash memory, a random access memory (RAM), a read only memory (ROM), or the like. The processing unit 22 creates initial setting information including date and time information for starting sensing, information indicating a time interval for measurement, and a subject ID. When the processing unit 22 detects that the sensor module 1 has started up and started up, the processing unit 22 creates initial setting information. The processing unit 22 outputs the created initial setting information to the first communication unit 21-1. The processing unit 22 acquires a sensor information acquisition request received by the second communication unit 21-2. The processing unit 22 creates a command for acquiring sensor information based on the acquired sensor information acquisition request. The processing unit 22 outputs the created command to the first communication unit 21-1. The processing unit 22 acquires the sensor information received by the first communication unit 21-1. The processing unit 22 acquires the sensor information request received by the second communication unit 21-2. The processing unit 22 creates a sensor information response including the acquired sensor information and addressed to the information processing device 3. The processing unit 22 outputs the created sensor information response to the second communication unit 21-2.

[0020] The processing unit 22 is realized, for example, by a hardware processor such as a CPU executing a computer program (software) stored in the storage unit 25. In addition, some or all of these functional units may be realized by hardware (including circuitry) such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a GPU (Graphics Processing Unit), or may be realized by cooperation between software and hardware. The computer program may be stored in advance in a storage device such as an HDD (Hard Disk Drive) or a flash memory, or may be stored in a removable storage medium such as a DVD or a CD-ROM, and installed by mounting the storage medium in a drive device.

[0021] (Information processing device 3) The information processing device 3 is realized by a device such as a personal computer, a server, a smartphone, a tablet computer, an industrial computer, etc. The information processing device 3 includes, for example, a communication unit 31, a processing unit 32, an operation unit 33, a derivation unit 34, a storage unit 35, and an output unit 36. The communication unit 31 is realized by a communication module. The communication unit 31 communicates with an external communication device via the network NW. The communication unit 31 may communicate using a communication method such as a wired LAN. The communication unit 31 holds communication information required to communicate with the data acquisition device 2 via the network NW. Specifically, the communication unit 31 transmits the sensor information acquisition request output by the processing unit 32 to the data acquisition device 2. The communication unit 31 transmits the sensor information request output by the processing unit 32 to the data acquisition device 2, and receives a sensor information response transmitted by the data acquisition device 2 in response to the transmitted sensor information request.

[0022] The processing unit 32 creates a sensor information acquisition request based on information output by the operation unit 33 requesting the data acquisition device 2 to acquire sensor information by operating the operation unit 33 as a result of the user operating the operation unit 33, and outputs the created sensor information acquisition request to the communication unit 31. The processing unit 32 creates a sensor information request based on information output by the operation unit 33 requesting the data acquisition device 2 to acquire sensor information by operating the operation unit 33 as a result of the user operating the operation unit 33, and outputs the created sensor information request to the communication unit 31. The processing unit 32 acquires a sensor information response transmitted by the data acquisition device 2 in response to the sensor information request from the communication unit 31. The processing unit 32 acquires date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, and digital data included in the acquired sensor information response. The processing unit 32 extracts digital data of the subject HS from the acquired date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, and digital data.

[0023] The operation unit 33 includes an input device and accepts operations. The input device includes a device for inputting character information such as a keyboard, a pointing device such as a mouse or a touch panel, a button, a dial, a joystick, a touch sensor, a touch pad, and the like. The user operates the operation unit 33 to input information requesting that the data acquisition device 2 acquire sensor information and a subject ID. The user operates the operation unit 33 to input information requesting that sensor information be acquired and a subject ID. The derivation unit 34 has a learning model 34a. The learning model 34a is obtained by machine learning the relationship between information specifying pressure and the flow rate corresponding to that pressure. When information specifying pressure is input, the learning model 34a outputs information specifying the flow rate. The derivation unit 34 acquires digital data (information specifying intranasal pressure) extracted by the processing unit 32. The derivation unit 34 inputs the acquired digital data (information specifying intranasal pressure) to the learning model 34a. The derivation unit 34 acquires information specifying the flow rate output by the learning model 34a for each of one or more pieces of information specifying the intranasal pressure. The output unit 36 ​​acquires date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, and the digital data from the processing unit 32. The output unit 36 ​​acquires information for identifying the flow rate corresponding to the acquired digital data (information for identifying the intranasal pressure) from the derivation unit 34. The output unit 36 ​​associates the acquired date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, the digital data, and information for identifying the flow rate, and stores them in the storage unit 35.

[0024] The storage unit 35 is realized by a HDD, a flash memory, a RAM, a ROM, or the like. The processing unit 32, the derivation unit 34, and the output unit 36 ​​are realized, for example, by a hardware processor such as a CPU executing a computer program (software) stored in the storage unit 35. In addition, some or all of these functional units may be realized by hardware (including circuitry) such as an LSI, ASIC, FPGA, or GPU, or may be realized by cooperation between software and hardware. The computer program may be stored in a storage device such as an HDD or flash memory in advance, or may be stored in a removable storage medium such as a DVD or CD-ROM, and installed by mounting the storage medium in a drive device.

[0025] Next, the operation of the sensing system 100 will be described. (Operation of sensing system 100) Fig. 4 is a flow diagram showing a first example of the operation of the sensing system of this embodiment. With reference to Fig. 4, the operation from when the data acquisition device 2 creates initial setting information to when the data acquisition device 2 receives sensor information from the sensor module 1 based on a sensor information acquisition request created by the information processing device 3 will be described. (Step S1-1) In the data acquisition device 2, the processing unit 22 creates initial setting information including date and time information for starting sensing, information indicating the time interval for measurement, and the subject ID. (Step S2-1) In the data acquisition device 2, the processing unit 22 outputs the created initial setting information to the first communication unit 21-1. The first communication unit 21-1 acquires the initial setting information output by the processing unit 22. The first communication unit 21-1 transmits the acquired initial setting information to the sensor module 1. (Step S3-1) In the sensor module 1 , the wireless transmitting / receiving unit 16 receives the initial setting information transmitted by the data acquisition device 2 . (Step S4-1) In the sensor module 1, the signal processing unit 14 acquires the initial setting information received by the wireless transceiver unit 16. The signal processing unit 14 acquires the date and time information for starting sensing, the information indicating the time interval of measurement, and the subject ID, which are included in the acquired initial setting information. The signal processing unit 14 stores the acquired date and time information for starting sensing, the information indicating the time interval of measurement, and the subject ID in the memory 15.

[0026] (Step S5-1) In the sensor module 1, the signal processing unit 14 sets, in the sensor 12, information indicating the time interval of the acquired measurement. (Step S6-1) In the sensor module 1, the sensor 12 acquires sensing data at a period corresponding to the measurement time interval. (Step S7-1) In the sensor module 1, the sensor 12 outputs acquired sensing data to the sensor module unit 13. The signal processing unit 14 acquires the sensing data output by the sensor 12 and converts the acquired sensing data into digital data. The signal processing unit 14 associates the digital data obtained by converting the sensing data with a measurement index and stores the associated data in the memory 15. The sensor module 1 associates digital data in daily life with a measurement index and stores the associated data in the memory 15 for a predetermined period of time. (Step S8-1) After a predetermined period of time has elapsed, the subject HS removes the sensor module 1. In the information processing device 3, the processing unit 32 creates a sensor information acquisition request based on an operation performed on the operation unit 33 by the user.

[0027] (Step S9-1) In the information processing device 3, the processing unit 32 outputs the created sensor information acquisition request to the communication unit 31. The communication unit 31 acquires the sensor information acquisition request output by the processing unit 32. The communication unit 31 transmits the acquired sensor information acquisition request to the data acquisition device 2. (Step S10-1) In the data acquisition device 2, the second communication unit 21-2 receives the sensor information acquisition request transmitted by the information processing device 3. (Step S11-1) In the data acquisition device 2, the processing unit 22 acquires the sensor information acquisition request received by the second communication unit 21-2. The processing unit 22 creates a command indicating acquisition of sensor information based on the acquired sensor information acquisition request. (Step S12-1) In the data acquisition device 2, the processing unit 22 outputs the created command to the first communication unit 21-1. The first communication unit 21-1 acquires the command output by the processing unit 22. The first communication unit transmits the acquired command to the sensor module 1.

[0028] (Step S13-1) In the sensor module 1 , the wireless transceiver 16 receives the command transmitted by the data acquisition device 2 . (Step S14-1) In the sensor module 1, the signal processing unit 14 acquires the command received by the wireless transceiver unit 16, and if the acquired command is to acquire sensor information, creates sensor information addressed to the data acquisition device 2, which includes information relating to the date and time information for starting sensing stored in the memory 15, information indicating the measurement time interval, the subject ID, the ID of the sensor module 1, the measurement index, and the digital data. (Step S15-1) In the sensor module 1, the signal processing unit 14 outputs the created sensor information to the wireless transceiver unit 16. The wireless transceiver unit 16 acquires the sensor information output by the signal processing unit 14, and transmits the acquired sensor information to the data acquisition device 2. (Step S16-1) In the data acquisition device 2, the first communication unit 21-1 receives the sensor information transmitted by the sensor module 1.

[0029] Fig. 5 is a flow diagram showing a second example of the operation of the sensing system of this embodiment. With reference to Fig. 5, the information processing device 3 creates a sensor information request, and acquires a sensor information response from the data acquisition device 2 based on the created sensor information request. The following describes the operation of the information processing device 3, from inputting digital data included in the acquired sensor information response to the learning model 34a, acquiring output data from the learning model 34a, and storing the acquired output data. (Step S1-2) In the information processing device 3, the processing unit 32 creates a sensor information request based on an operation performed on the operation unit 33 by the user. (Step S2-2) In the information processing device 3, the processing unit 32 outputs the created sensor information request to the communication unit 31. The communication unit 31 acquires the sensor information request output by the processing unit 32, and transmits the acquired sensor information request to the data acquisition device 2. (Step S3-2) In the data acquisition device 2, the second communication unit 21-2 receives the sensor information request transmitted by the information processing device 3. (Step S4-2) In the data acquisition device 2, the processing unit 22 acquires the sensor information request received by the second communication unit 21-2. The processing unit 22 creates a sensor information response based on the acquired sensor information request. (Step S5-2) In the data acquisition device 2, the processing unit 22 outputs the created sensor information response to the second communication unit 21-2. The second communication unit 21-2 acquires the sensor information response output by the processing unit 22. The second communication unit 21-2 transmits the acquired sensor information response to the information processing device 3.

[0030] (Step S6-2) In the information processing device 3, the communication unit 31 receives the sensor information response transmitted by the data acquisition device 2. (Step S7-2) In the information processing device 3, the processing unit 32 acquires the sensor information response received by the communication unit 31. The processing unit 32 acquires date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, and digital data, which are included in the acquired sensor information response. The processing unit 32 extracts digital data of the subject HS from the acquired date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, and digital data. (Step S8-2) In the information processing device 3, the derivation unit 34 acquires the digital data of the subject HS extracted by the processing unit 32. The derivation unit 34 inputs the acquired digital data of the subject HS to the learning model 34a. The derivation unit 34 acquires output data from the learning model 34a for the digital data of the subject HS. (Step S9-2) In the information processing device 3, the output unit 36 ​​acquires date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, and digital data from the processing unit 32. The output unit 36 ​​acquires information for identifying the flow rate corresponding to the acquired digital data (information for identifying the intranasal pressure) from the derivation unit 34. The output unit 36 ​​associates the acquired date and time information for starting sensing, information indicating the time interval of measurement, the subject ID, the ID of the sensor module 1, the measurement index, the digital data, and information for identifying the flow rate, and stores them in the storage unit 35.

[0031] In the above embodiment, the data acquisition device 2 and the information processing device 3 are connected via a network NW, but the present invention is not limited to this example. For example, the data acquisition device 2 may be connected to the information processing device 3 via a wired connection. By connecting the data acquisition device 2 and the information processing device 3 via the network NW, even if the subject HS goes to a remote location, the user can access the digital data of the subject HS. In the above-described embodiment, the sensor module 1 and the data acquisition device 2 are wirelessly connected, but the present invention is not limited to this example. For example, the sensor module 1 and the data acquisition device 2 may be wired. In this case, the sensor module unit 13 includes a module for wired transmission and reception instead of the wireless transmission and reception unit 16. In the above embodiment, the sensing system 100 includes one sensor module 1, but the present invention is not limited to this example. For example, the sensing system 100 may include a plurality of sensor modules.

[0032] In the above-mentioned sensing system 100, the sensor module 1 may further include at least one of a temperature sensor, a humidity sensor, an acceleration sensor, and a pulse wave sensor. In this case, the learning model 34a is created by machine learning the relationship between at least one of information specifying the temperature, information specifying the humidity, information specifying the acceleration, and information specifying the pulse wave, and information specifying the flow rate, in addition to information specifying the pressure. By configuring in this way, it is possible to combine biological information sensed by other sensors, so that the environment in which the fluid is placed can be understood. Therefore, it is possible to construct a learning model that is robust against measurement noise, posture noise, and the like.

[0033] According to the sensing system 100 of this embodiment, the sensing system 100 senses biological information using a sensor module. The sensing system 100 includes a sensor module 1 that stores pressure data acquired by sensing pressure occurring at a predetermined part of a subject, and an information processing device 3 that acquires the pressure data stored by the sensor module 1 and the subject's identification information, and derives flow data from the acquired pressure data for each of the subject's identification information using a learning model 34a created by machine learning the relationship between information specifying pressure and information specifying flow rate. With this configuration, the pressure data can be converted into flow data using a learning model 34a created by machine learning the relationship between information specifying the pressure and information specifying the flow rate, thereby improving the accuracy of estimating the flow rate. Machine learning includes LightGBM, which is a machine learning framework of gradient boosting based on a decision tree algorithm, neural networks such as 1DCNN, CNN, LSTM, and Transformer, regression analysis (multiple regression analysis, etc.) which is a statistical learning method, and LightGBM of gradient boosting. Since various learning models can be used, it is possible to provide flexibility in the size (number of parameters) and accuracy of the model, estimation of secondary state quantities, etc.

[0034] In the above-described sensing system 100, the sensor module 1 is installed in the nasal cavity. With this configuration, the sensing system 100 can derive information specifying the flow rate from information specifying the pressure in the nasal cavity using the learning model 34a, and can therefore derive information specifying the flow rate of exhaled or inhaled air flowing through the nasal cavity. In the above-mentioned sensing system 100, the sensor module 1 comprises a sensor 12 that senses pressure, a battery 11 that supplies power, a signal processing unit 14 that creates digital data based on the results of sensing by the sensor 12, a memory 15 that stores the digital data created by the signal processing unit 14, and a wireless transceiver unit 16 that receives a command to acquire data sent by the data acquisition device 2 and transmits the digital data stored in the memory 15 and the subject's identification information to the data acquisition device 2 based on the received command.

[0035] By providing the sensor module 1 with the wireless transceiver 16, sensor information can be transmitted to the outside without providing the sensor module 1 with a connector for reading out digital data. By providing the sensor module 1 with the memory 15, digital data sensed by the sensor 12 can be stored. Furthermore, when the sensor module 1 is removed from the nasal cavity, the digital data stored in the memory 15 can be transmitted to the outside. This makes it possible to acquire time-series data with no gaps in time. If the sensor module 1 did not have the memory 15, the digital data sensed by the sensor 12 would need to be wirelessly transmitted to the outside without being stored. However, since the sensor module 1 may be installed in the nasal cavity, radio waves would have difficulty leaving the nasal cavity, and there is a risk that data will not be transmitted. Furthermore, transmitting data wirelessly every time sensing is performed increases battery energy consumption. According to the sensing system 100 of this embodiment, when the sensor module 1 is removed from the nasal cavity, the digital data stored in the memory 15 can be transmitted wirelessly, so battery energy consumption is reduced. This makes it possible to extend the life of the battery energy.

[0036] In the above-described sensing system 100, the sensor 12 senses either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity. The information processing device 3 uses the learning model 34a to derive either or both of the flow rate in the left nasal cavity and the flow rate in the right nasal cavity from either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity. By configuring in this manner, the sensing system 100 can derive either or both of information specifying the flow rate in the left nasal cavity and information specifying the flow rate in the left nasal cavity using the learning model 34a from either or both of information specifying the pressure in the left nasal cavity and information specifying the pressure in the right nasal cavity, and can therefore derive either or both of information specifying the flow rate of exhaled or inhaled air flowing through the left nasal cavity and information specifying the flow rate of exhaled or inhaled air flowing through the right nasal cavity.

[0037] In the above-mentioned sensing system 100, the sensor module 1 further includes at least one of a temperature sensor, a humidity sensor, an acceleration sensor, and a pulse wave sensor, and the learning model is created by machine learning the relationship between the information specifying the pressure and the information specifying the flow rate based on at least one of the information specifying the temperature, the information specifying the humidity, the information specifying the acceleration, and the information specifying the pulse wave. By configuring in this way, it is possible to combine biological information sensed by other sensors, so that it is possible to grasp the environment in which the fluid is placed. Therefore, it is possible to construct a learning model that is robust against measurement noise, posture noise, and the like.

[0038] Next, a method for creating the learning model 34a will be described. The learning model 34a is created by the learning device 4. The learning device 4 will be described. (Learning device 4) 6 is a diagram showing an example of a learning device according to the present embodiment. The learning device 4 acquires one or more pieces of information specifying pressure and information specifying a flow rate at which each of the one or more pressures is obtained. The learning device 4 creates a learning model 34a by machine learning the relationship between the information specifying pressure and the information specifying flow rate based on the acquired one or more pieces of information specifying pressure and the information specifying a flow rate at which each of the one or more pressures is obtained. The learning device 4 includes an input unit 41, a processing unit 42, a storage unit 45, and an output unit 46.

[0039] Input data is input to the input unit 41. An example of the input data is information specifying pressures and information specifying flow rates at which the respective pressures are obtained. Specifically, an example of the pressures and flow rates is information obtained simultaneously by a pressure sensor and a flow rate sensor. FIG. 7 is a diagram showing Example 1 of input data. In FIG. 7, the horizontal axis is elapsed time [sec], and the vertical axis is measured values ​​([units: Press [Pa] and Flow rate [slm]). An example of input data is sampled at a predetermined period such as 0.05 seconds. In FIG. 7, the solid line is a time series waveform of the measured pressure data, and the dashed line is a time series waveform of the measured flow rate data. Fig. 8 is a diagram showing Example 2 of input data. Fig. 8 shows a scatter plot of the input data. In Fig. 8, the horizontal axis is flow rate [slm] and the vertical axis is pressure [Pa]. Fig. 8 shows that there is a positive correlation between the flow rate and the pressure. Fig. 9 is a diagram showing input data example 3. Fig. 9 shows Pearson's correlation coefficient as an example of a correlation map of input data.

[0040] 7 to 9, it can be seen that there is a linear relationship between pressure and flow rate. Also, as shown in Fig. 7, the time series waveforms of the pressure data and the flow rate data have characteristic portions, such as a section where the flow rate decreases significantly after a positive peak and then the slope becomes gentle. Specifically, this characteristic portion can be seen in Fig. 7 at elapsed times of 20 seconds, 26 seconds, and 32 seconds. This characteristic portion corresponds to the section (time) of one breath. If the flow rate were to be obtained from the pressure through a simple linear transformation, it would not be possible to reproduce this characteristic waveform. This is because a simple linear transformation can only express magnitude relationships, and the flow rate data and pressure data used as a reference do not behave in exactly the same way. It can be seen that there is a strong correlation between pressure and flow rate (Pearson's correlation coefficient: 0.924). Let's return to Figure 6 and continue our explanation.

[0041] The processing unit 42 acquires one or more pieces of input data input to the input unit 41. The processing unit 42 creates a learning model 34a by machine learning the relationship between the information specifying the pressure and the information specifying the flow rate based on the information specifying the pressure included in each of the acquired one or more pieces of input data and the information specifying the flow rate at which that pressure is obtained. When machine learning the relationship between the information specifying the pressure and the information specifying the flow rate, the processing unit 42 machine-learns the relationship between the waveform of the pressure and the waveform of the flow rate. With this configuration, it is possible to create a learning model 34a that converts the information specifying the pressure into a flow rate based on the characteristics of the waveform in addition to the magnitude of the pressure. In this embodiment, as an example, a case where a one-dimensional convolutional neural network (1DCNN) is used for machine learning of the relationship between information specifying the pressure and information specifying the flow rate will be described.

[0042] 10 is a diagram for explaining an example of the processing of the learning device according to the present embodiment. The processing of the processing unit 42 will be explained using an encoder-decoder model. An example of the input data is sampled every predetermined period (section), such as 0.05 seconds. The processing unit 42 blocks every predetermined section, such as at least one breathing section (time), from one or more pieces of information specifying the pressure and one or more pieces of information specifying the flow rate sampled every predetermined period (section). Here, between a block (e.g., block BL01) and an adjacent block (e.g., block BL02) that is shifted by a certain time from the block (e.g., block BL01) (between adjacent blocks), there may be an overlap between the information specifying the pressure and the information specifying the flow rate. The processing unit 42 acquires the information specifying the pressure and the information specifying the flow rate for each block based on the blocking result. By configuring in this way, multiple blocks can be created from one time series waveform, so that the amount of data used for machine learning can be increased.

[0043] The processing unit 42 inputs the information specifying the blocked pressure to an encoder-decoder model, compresses it using 1DCNN, and restores it to the same size. For example, the processing unit 42 inputs the information specifying the blocked pressure (time series data) to an encoder. The encoder generates a fixed-length vector based on the information specifying the blocked pressure (time series data). The processing unit 42 inputs the fixed-length vector output by the encoder to a decoder. The decoder generates time series data based on the fixed-length vector. The processing unit 42 creates a learning model by machine learning an encoder-decoder model so that the error between the information specifying the flow rate included in the same block as the result of restoring to the same size is small. For example, the processing unit 42 creates a learning model by machine learning an encoder-decoder model so that the error between the time series data generated by the decoder and the information specifying the flow rate included in the block including the time series data input to the encoder is small. The output unit 46 acquires the learning model created by the processing unit 42. The output unit 46 outputs the acquired learning model. Specifically, the output unit 46 may store the learning model in the storage unit 45, or may output the learning model to the outside of the learning device 4. When the learning model is output to the outside of the learning device 4, it may be output to the information processing device 3. In this case, in the information processing device 3, the communication unit 31 receives the learning model output by the learning device 4. The processing unit 32 acquires the learning model received by the communication unit 31. The processing unit 32 outputs the acquired learning model to the derivation unit 34. The derivation unit 34 acquires the learning model output by the processing unit 32. The derivation unit 34 sets the acquired learning model as learning model 34a. In the information processing device 3, the derivation unit 34 inputs information specifying the pressure to the learning model 34a. When the information specifying the pressure is input, the learning model 34a outputs information specifying the flow rate.

[0044] FIG. 11 is a diagram showing an example of output data of the information processing device according to this embodiment. In FIG. 11, the solid line indicates information specifying the flow rate. This information specifying the flow rate is acquired from the data acquisition device 2, and is validation data output by the flowmeter. The dashed line indicates information specifying the flow rate. This information specifying the flow rate is output by the learning model 34a. Here, the learning model 34a created by machine learning using an encoder-decoder model is used. According to FIG. 11, the estimation accuracy of the information specifying the flow rate output by the learning model 34a is approximately 4.5 slm based on the squared error. 2 From this estimation accuracy, the information specifying the flow rate obtained by the information processing device 3 from the information specifying the pressure is effective.

[0045] (Operation of learning device 4) FIG. 12 is a diagram showing an example of the operation of the learning device according to this embodiment. (Step S1-3) In the learning device 4, the input unit 41 acquires input data. An example of the input data includes information specifying a pressure and information specifying a flow rate. An example of the pressure and the flow rate is information acquired simultaneously by a pressure sensor and a flow rate sensor. (Step S2-3) In the learning device 4, the processing unit 42 acquires input data input to the input unit 41. The processing unit 42 creates a learning model 34a by machine learning the relationship between the information specifying the pressure and the information specifying the flow rate, based on the information specifying the pressure included in the acquired input data and the information specifying the flow rate at which that pressure is obtained. (Step S3-3) In the learning device 4, the output unit 46 acquires the learning model created by the processing unit 42. The output unit 46 outputs the acquired learning model.

[0046] In the above-mentioned embodiment, as an example, the case where the learning model 34a is created using 1DCNN has been described, but the present invention is not limited to this example. For example, machine learning models such as a support vector machine (SVM), a LightGBM (Light Gradient Boosting Machine), a Long short-term memory (LSTM), a Transformer, a fully connected neural network other than a CNN (Convolutional Neural Network), regression analysis (multiple regression analysis, etc.) which is a statistical learning method, and LightGBM of gradient boosting can be applied. LightGBM is a machine learning framework of gradient boosting based on a decision tree algorithm. LSTM is used for seq2seq in natural language, etc. However, if the model does not take into account information before and after the data as a time series, such as LSTM or 1DCNN, it is necessary to design separate features so that they can be taken into account.

[0047] Moreover, the output of the learning model 34a may be one-dimensional or two or more dimensional. By making it two or more dimensional, information specifying a plurality of flow rates can be estimated at the same time. In the above-described embodiment, in the learning device 4, the input unit 41 may input at least one of information specifying temperature, information specifying humidity, information specifying acceleration, and information specifying pulse wave in addition to information specifying pressure. Here, an example of at least one of information specifying temperature, information specifying humidity, information specifying acceleration, and information specifying pulse wave is acquired together with information specifying pressure and information specifying flow rate. The processing unit 42 creates a learning model by machine learning the relationship between at least one of information specifying temperature, information specifying humidity, information specifying acceleration, and information specifying pulse wave and information specifying flow rate in addition to information specifying pressure. By configuring in this way, it is possible to combine biological information sensed by other sensors, so that it is possible to grasp the environment in which the fluid is placed. Therefore, it is possible to construct a learning model that is robust against measurement noise, posture noise, and the like.

[0048] The learning device 4 of this embodiment includes an input unit 41 that receives information specifying pressure and information specifying a flow rate at which the pressure is obtained, a processing unit 42 that creates a learning model by machine learning the relationship between the information specifying pressure and the information specifying flow rate based on the information specifying pressure and the information specifying flow rate received by the input unit 41, and an output unit 46 that outputs the learning model created by the processing unit 42. With this configuration, by reflecting the information specifying the fluid that appears as pressure as a feature of machine learning, a simple one-to-one correspondence between pressure and flow rate, such as in Bernoulli's theorem, can be made many-to-one as in the learning device 4, thereby improving the estimation accuracy of the flow rate.

[0049] In the above-mentioned learning device 4, the information specifying the pressure is the pressure in the nasal cavity during at least one breath, and the information specifying the flow rate is the flow rate in the nasal cavity during at least one breath. By configuring in this way, machine learning can be performed based on the characteristic parts of the time-series waveform of the pressure data. In the above-mentioned learning device 4, the information specifying the pressure is either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity, and the information specifying the flow rate is either or both of the flow rate in the left nasal cavity and the flow rate in the right nasal cavity. By configuring in this way, it is possible to machine-learn the relationship between either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity and either or both of the flow rate in the left nasal cavity and the flow rate in the right nasal cavity based on either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity and either or both of the flow rate in the left nasal cavity.

[0050] Although the embodiments of the present invention have been described above, these embodiments are presented as examples and are not intended to limit the scope of the invention. The embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the gist of the invention. The embodiments and their modifications include, for example, those that can be easily imagined by a person skilled in the art, those that are substantially the same, those that are within the scope of the equivalents, and the like. For example, a computer program for implementing the functions of each of the above-mentioned devices may be recorded on a computer-readable recording medium, and the computer program recorded on the recording medium may be read into a computer system and executed. Note that the "computer system" referred to here may include hardware such as an OS and peripheral devices. In addition, the term "computer-readable recording medium" refers to writable non-volatile memory such as a flexible disk, an optical magnetic disk, a ROM, or a flash memory, a portable medium such as a DVD (Digital Versatile Disc), or a storage device such as a hard disk built into a computer system. Furthermore, the term "computer-readable recording medium" includes devices that retain a program for a certain period of time, such as volatile memory (e.g., Dynamic Random Access Memory (DRAM)) within a computer system that serves as a server or client when a computer program is transmitted via a network such as the Internet or a communication line such as a telephone line. The program may be transmitted from a computer system in which the program is stored in a storage device or the like to another computer system via a transmission medium, or by a transmission wave in the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing part of the functions described above. Furthermore, the above-mentioned functions may be realized in combination with a program already recorded in the computer system, that is, a so-called differential file (differential program). [Explanation of symbols]

[0051] 100...sensing system, 1...sensor module, 2...data acquisition device, 3...information processing device, 11...battery, 12...sensor, 13...sensor module section, 14...signal processing section, 15...memory, 16...wireless transmission / reception section, 21-1...first communication section, 21-2...second communication section, 22...processing section, 25...storage section, 31...communication section, 32...processing section, 33...operation section, 34...derivation section, 34a...learning model, 35...storage section, 36...output section

Claims

1. A sensing system that senses biological information using a sensor module, a sensor module that stores pressure data acquired by sensing pressure generated at a predetermined part of the subject; an information processing device that acquires the pressure data and identification information of the subject stored in the sensor module, and derives flow data from the acquired pressure data for each identification information of the subject using a learning model created by machine learning the relationship between pressure information that is information specifying a pressure and flow information that is information specifying a flow rate; Equipped with The sensor module is placed in the nasal cavity, The learning model is created by dividing the pressure information and the flow information sampled for each predetermined section into blocks for at least one breathing section, and learning a relationship between the pressure information and the flow information for each block based on the blocking result, A sensing system having information overlap, in which the pressure information and the flow rate information at a certain time are included in both a first block and a second block that are adjacent to each other in the time axis direction among the blocks.

2. A sensing system that senses biological information using a sensor module, a sensor module that stores pressure data acquired by sensing pressure generated at a predetermined part of the subject; an information processing device that acquires the pressure data and identification information of the subject stored in the sensor module, and derives flow data from the acquired pressure data for each identification information of the subject using a learning model created by machine learning the relationship between pressure information that is information specifying a pressure and flow information that is information specifying a flow rate; Equipped with The sensor module is placed in the nasal cavity, A sensing system, wherein the pressure includes pressure data of static pressure only.

3. The sensor module includes: A sensor for sensing pressure; A battery for supplying power; A signal processing unit that creates digital data based on the result of sensing by the sensor; A memory for storing the digital data created by the signal processing unit; a wireless transceiver that receives a command to acquire data transmitted by an acquisition device, and transmits the digital data stored in the memory and the identification information of the subject to the acquisition device based on the received command; The sensing system according to claim 1 or 2, comprising:

4. The sensor senses either or both of a pressure in the left nasal cavity and a pressure in the right nasal cavity; The sensing system of claim 3, wherein the information processing device uses the learning model to derive either or both of the flow rate in the left nasal cavity and the flow rate in the right nasal cavity from either or both of the pressure in the left nasal cavity and the pressure in the right nasal cavity.

5. The sensor module further includes at least one of a temperature sensor, a humidity sensor, an acceleration sensor, and a pulse wave sensor; The learning model is created by machine learning the relationship between the pressure information and the flow rate information based on at least one of information specifying temperature, information specifying humidity, information specifying acceleration, and information specifying a pulse wave. The sensing system according to any one of claims 1 to 3.

6. A sensing method performed by a sensing system that senses biological information using a sensor module, comprising: A step of storing pressure data acquired by sensing the pressure occurring at a predetermined part of the subject; acquiring the pressure data and the identification information of the subject stored in the sensor module, and deriving flow data from the acquired pressure data for each identification information of the subject using a learning model created by machine learning the relationship between pressure information, which is information specifying a pressure, and flow information, which is information specifying a flow rate; having The sensor module is placed in the nasal cavity, The learning model is created by dividing the pressure information and the flow information sampled for each predetermined section into blocks for at least one breathing section, and learning a relationship between the pressure information and the flow information for each block based on the blocking result, The pressure information and the flow rate information at a certain time are included in both a first block and a second block that are adjacent to each other in the time axis direction among the blocks. Sensing method.

7. A sensing method performed by a sensing system that senses biological information using a sensor module, comprising: A step of storing pressure data acquired by sensing the pressure occurring at a predetermined part of the subject; acquiring the pressure data and the identification information of the subject stored in the sensor module, and deriving flow data from the acquired pressure data for each identification information of the subject using a learning model created by machine learning the relationship between pressure information, which is information specifying a pressure, and flow information, which is information specifying a flow rate; having The sensor module is placed in the nasal cavity, The pressure includes static pressure only. Sensing method.

Images