Biological component measurement device and biological component measurement method

Abstract

In the present invention, an excitation-light source (1) emits excitation light that advances through an optical medium (3) toward a sample (5) placed on a sample placement surface. A probe light source (2) emits probe light that advances through the optical medium (3). An optical location detector (4) detects the location of emitted probe light that has been emitted from the optical medium (3). A calculator (11) calculates the amount or the concentration of a biological component in a biological substance, on the basis of: the location of the emitted probe light detected by the optical location detector (4) when the sample (5) is a reference substance that absorbs the excitation light and that does not have a specific absorption peak within the excitation wavelength of the excitation light; and the location of the emitted probe light detected by the optical location detector (4) when the sample (5) is the biological substance.

Description

Biological component measuring device and biological component measuring method

[0001] The present disclosure relates to a biological component measuring device and a biological component measuring method.

[0002] Conventionally, biological component measuring devices and biological component measuring methods have been known. For example, the method described in JP 2017-519214 A (Patent Document 1) includes the steps of: disposing an optical medium on the surface of a substance so that at least one region of the surface of the optical medium is in contact with the substance surface; irradiating the substance surface with an excitation light beam having an excitation wavelength through the surface region of the optical medium in contact with the substance surface; emitting a probe light beam through the optical medium onto a surface region of the optical medium in direct contact with the substance surface, the probe light beam being emitted through the optical medium so that the probe light beam and the excitation light beam overlap at an interface between the optical medium and the substance surface and the probe light beam is reflected at the interface; directly or indirectly detecting a polarization of the reflected probe light beam in accordance with the wavelength of the excitation light beam; and analyzing the substance based on the polarization of the reflected probe light beam in accordance with the wavelength of the excitation light beam.

[0003] Special table 2017-519214 publication

[0004] The device disclosed in Patent Document 1 uses an infrared light source to emit an excitation light beam. The infrared light source is affected by external factors such as heat accumulation inside the light source or changes in the external temperature, resulting in unstable infrared light output. As a result, the heat generated by biological materials due to irradiation with infrared light also fluctuates, resulting in unstable measurement accuracy of the amount or concentration of biological components.

[0005] Therefore, an object of the present disclosure is to provide a biological component measuring device and a biological component measuring method that can accurately measure the amount or concentration of a biological component without being affected by the instability of infrared light.

[0006] The biological component measuring device of the present disclosure includes an optical medium including a sample mounting surface, an excitation light source that travels through the optical medium toward a sample placed on the sample mounting surface and emits excitation light having an excitation wavelength of the sample, a probe light source that emits probe light that travels through the optical medium, a light position detector that detects the position of the emitted probe light emitted from the optical medium, and a calculator that calculates the amount or concentration of the biological component in the biological material based on the position of the emitted probe light detected by the light position detector when the sample is a reference material and the position of the emitted probe light detected by the light position detector when the sample is a biological material. The reference material absorbs the excitation light but does not have a specific absorption peak within the excitation wavelength of the excitation light.

[0007] The disclosed method for measuring a biological component includes the steps of: an excitation light source emitting excitation light having an excitation wavelength of the sample, traveling through the optical medium toward a sample placed on a sample-mounting surface of the optical medium; a probe light source emitting probe light traveling through the optical medium; a light position detector detecting the position of the outgoing probe light of the probe light incident on the optical medium when the sample is a reference substance; a light position detector detecting the position of the outgoing probe light of the probe light incident on the optical medium when the sample is a biological substance; and a computing unit calculating the amount or concentration of the biological component in the biological substance based on the position of the outgoing probe light detected by the light position detector when the sample is the reference substance and the position of the outgoing probe light detected by the light position detector when the sample is a biological substance. The reference substance absorbs the excitation light but does not have a specific absorption peak within the excitation wavelength of the excitation light.

[0008] According to the present disclosure, the amount or concentration of a biological component in a biological material is calculated based on the position of the emitted probe light detected by the optical position detector when the sample is a reference material and the position of the emitted probe light detected by the optical position detector when the sample is a biological material, thereby enabling the amount or concentration of the biological component to be measured accurately without being affected by the instability of infrared light.

[0009] 1 is a diagram showing the configuration of a biological component measuring device according to embodiment 1. (a) and (b) are diagrams showing the optical path of probe light 7. (b) is a flowchart showing the procedure of a biological component measurement method by the biological component measuring device according to embodiment 1. (c) is a diagram showing the configuration of a biological component analyzer according to a modified example of embodiment 1. (d) is a flowchart showing the procedure of a biological component measurement method by the biological component measuring device according to embodiment 2. (e) is a diagram showing the configuration of a computing unit 11. (f) is a diagram showing the configuration of a learning device 91 according to embodiment 3. (g) is a diagram showing the configuration of an inference device 95 according to embodiment 3. (f) is a diagram showing the configuration of a learning device 91A according to embodiment 4. (g) is a diagram showing the configuration of an inference device 95A according to embodiment 4. (f) is a diagram showing the configuration of a learning device 91B according to embodiment 5. (f) is a diagram showing the configuration of an inference device 95B according to embodiment 5. (f) is a diagram showing the configuration of a learning device 91C according to embodiment 6. (f) is a diagram showing the configuration of an inference device 95C according to embodiment 6. FIG. 10 is a diagram showing a configuration in which the function of the arithmetic unit 11 is realized using software.

[0010] 1 is a diagram showing the configuration of a biological component measuring device according to embodiment 1. The biological component measuring device includes an excitation light source 1, a probe light source 2, an optical medium 3, a light position detector 4, and a computing unit 11.

[0011] The excitation light source 1 includes at least one infrared light source. The excitation light source 1 includes a broadband quantum cascade laser that emits infrared light in all wavelength ranges including the excitation wavelengths (wavelengths of the fingerprint spectrum) of the component to be measured in the sample 5, or in a portion of the wavelength range. When the component to be measured is sugar in the human body, the wavelengths used for measurement are, for example, λ1, λ2, and λ3. The light with wavelengths λ1 and λ2 are excitation wavelengths of sugar in the human body and are absorbed by sugar in the human body. The light with wavelengths λ3 is not absorbed by sugar in the human body and is used as a reference wavelength that is absorbed by a reference substance. Four or more wavelengths may be used for measurement.

[0012] Infrared light emitted from the excitation light source 1 passes through the optical medium 3 as excitation light 6 and enters the sample 5. When measuring components contained in a living body, the excitation light 6 is incident on the skin of the living body's finger, arm, ear, etc., and the absorption of light by substances contained inside the living body, such as the absorption of substances contained in interstitial fluid, is measured.

[0013] The probe light source 2 emits probe light 7. The probe light 7 enters the optical medium 3 from the third surface 33 of the optical medium 3. The probe light 7 is refracted at the third surface 33 and travels through the optical medium 3 toward the interface between the optical medium 3 (second surface 32) and the sample 5.

[0014] 2(a) and 2(b) are diagrams showing the optical path of the probe light 7. In a plan view of the sample placement surface (second surface 32), the optical path of the probe light 7 in the optical medium 3 partially (FIG. 2(a)) or entirely (FIG. 2(b)) overlaps with the portion of the sample placement surface (second surface 32) that is irradiated by the excitation light 6.

[0015] The probe light 7 is totally internally reflected at the interface between the optical medium 3 (second surface 32) and the sample 5. While the probe light 7 travels through the optical medium 3, the probe light 7 travels through a refractive index gradient region 8 that is generated in the optical medium 3 by heat absorbed by the sample 5. The probe light 7 is refracted in the refractive index gradient region 8, and the traveling direction of the probe light 7 is changed. The probe light 7 is emitted from the fourth surface 34 of the optical medium 3.

[0016] The wavelength of the light output from the probe light source 2 may be in any wavelength band that is transmitted through the optical medium 3. For example, a wavelength band such as the visible light range of 400 to 900 nm, which is mass-produced for many purposes and is inexpensive, or the invisible wavelength band of 1300 to 1700 nm, which is used in optical fiber communications, may be used.

[0017] The optical medium 3 is made of a material that transmits the excitation light 6 emitted from the excitation light source 1 and the incident probe light 7 emitted from the probe light source 2. For example, the optical medium 3 is made of a substance such as zinc sulfide (ZnS) or zinc selenide (ZnSe), which generally has high transmittance in the wavelength range from visible light to infrared light. When the measurement target is a living body, zinc sulfide (ZnS) can be used as the optical medium 3 in consideration of biological stability. Chalcogenide glass, which has a lower thermal conductivity than zinc sulfide (ZnS) or zinc selenide (ZnSe), may be used so that changes in the refractive index of the optical medium 3 due to heat generated in the sample 5 are localized.

[0018] A case where the optical output of the excitation light source 1 is zero (reference state) will be described. In the reference state, the internal state of the optical medium 3 is uniform, so the light output from the probe light source 2 is refracted only when it enters and exits the optical medium 3. In the reference state, the position where the emitted probe light 7a enters the light position detector 4 is defined as the reference position RP. In this embodiment, the path of the probe light 7a is shown as a path that undergoes one total reflection at the interface between the optical medium 3 and the sample 5. However, any path that passes through a refractive index gradient region 8 generated in the optical medium 3 may be used, and the path may also be a path that undergoes two or more total reflections within the optical medium, or a path that passes near the contact surface with the sample 5 parallel to the contact surface.

[0019] Next, a case where there is output from the excitation light source 1 will be described. The excitation light source 1 outputs infrared light of the fingerprint spectrum wavelength of the component to be measured in the sample 5 as excitation light 6. The excitation light 6 is incident on the sample 5 via the optical medium 3. The excitation light 6 is absorbed by the sample 5. Absorption heat is generated in the sample 5 due to the absorption. The generated absorption heat propagates into the optical medium 3, causing a temperature gradient inside the optical medium 3. Since the refractive index of an optical medium generally has temperature dependency, a refractive index gradient region 8 is formed in accordance with the temperature gradient. This state is called a changed state.

[0020] In the change state, the incident probe light 7 output from the probe light source 2 passes through the refractive index gradient region 8. The incident probe light 7 is refracted in accordance with the gradient of the refractive index at the position in the refractive index gradient region 8 through which it passes. The refracted incident probe light 7 is emitted from the optical medium 3 as emitted probe light 7b and enters the optical position detector 4. In the change state, the position at which the emitted probe light 7b enters the optical position detector 4 is designated as CP. Although FIG. 1 shows the incident position only in the height direction, the incident position also displaces in the horizontal direction. The incident positions of the excitation light 6 and the incident probe light 7 are adjusted to positions where the difference between the reference position RP and the position CP is maximized, i.e., where the refractive index gradient is maximized.

[0021] If the amount of the component to be measured in the sample 5 (i.e., the component with a high absorption coefficient at a certain wavelength of light emitted from the excitation light source 1) is large, the amount of light absorbed by the sample 5 increases, and therefore the amount of heat generated increases. As a result, the refractive index gradient increases. The relationship between the difference between the reference position RP at which the emitted probe light 7a is incident on the light position detector 4 in the reference state and the position CP at which the emitted probe light 7b is incident on the light position detector 4 in the changed state and the amount of the component to be measured in the sample 5 is approximately proportional.

[0022] The excitation light source 1 inherently has a fixed output for each wavelength. However, the excitation light source 1 is affected by external factors such as heat accumulation inside the excitation light source 1 and changes in external temperature. In measuring biological materials, the position CP of the emitted probe light 7b is affected by external factors. In this embodiment, the output reflecting the influence of these external factors is measured using absorption spectroscopy with a reference substance. Since absorption spectroscopy uses the ease with which a substance absorbs light of a given wavelength (absorption coefficient), a substance with a constant absorption coefficient (no unique absorption peak) in the wavelength band of the excitation light source 1 is appropriate as the reference substance. Measurement using such a reference substance as the sample 5 is equivalent to measuring the output of excitation light corresponding to the absorption coefficient of that substance.

[0023] In this embodiment, the influence of external factors is eliminated using not only the position CP(B) of the emitted probe light 7b(B) when the biological material is used as the sample 5, but also the position CP(A) of the emitted probe light 7b(A) when the reference material is used as the sample 5. The position CP(A) serves as an evaluation index for the stability of the optical output of the excitation light 6 during measurement. By detecting the position CP(A) when the sample 5 and the reference material are used before measuring the amount or concentration of a biological component in the biological material, it is possible to perform measurements that eliminate the influence of external factors of the excitation light source 1. As the reference material, a material that absorbs the wavelength band of the excitation light source 1, does not have a specific absorption peak in that wavelength band, and is in close contact with the optical medium 3 is selected. For example, in the infrared light region, water is an example. The refractive index gradient formed when the reference material is the sample 5 is fundamentally different from the refractive index gradient formed when the biological material is used as the sample 5.

[0024] The calculator 11 calculates the output from the light position detector 4. The calculator 11 measures the amount or concentration of a biological component (such as sugar) in the biological material based on the position CP(A) of the emitted probe light 7b detected by the light position detector 4 when the sample 5 is a reference material and the position CP(B) of the emitted probe light 7b detected by the light position detector 4 when the sample 5 is a biological material. For example, the calculator 11 calculates CP(B)-CP(A) or CP(B) / CP(A) to determine the amount or concentration of a component of the biological material (such as sugar in the skin).

[0025] FIG. 3 is a flowchart showing the steps of the method for measuring a biological component by the biological component measuring device according to the first embodiment.

[0026] In step S101, the sample 5 is set to a reference substance (e.g., water). In step S102, in a changing state in which the excitation light 6 from the excitation light source 1 is irradiated onto the sample 5, the light position detector 4 detects the position CP(A) of the emitted probe light 7b(A).

[0027] In step S103, the sample 5 is set on a biological material (for example, skin). In step S104, in a changing state in which the excitation light 6 from the excitation light source 1 is irradiated onto the sample 5, the light position detector 4 detects the position CP(B) of the emitted probe light 7b(B).

[0028] In step S105, the calculator 11 calculates the amount or concentration of a component of the biological material (for example, sugar in the skin) based on the positions CP(A) and CP(B). For example, the calculator 11 calculates CP(B)-CP(A) or CP(B) / CP(A) to determine the amount or concentration of the component of the biological material (for example, sugar in the skin).

[0029] According to this embodiment, it is possible to monitor the output of excitation light that is affected by external factors without adding any new components by using the measurement results when the reference substance is used as sample 5. By measuring the reference substance and the biological substance together, it is possible to eliminate the instability of the excitation light, and therefore it is possible to measure the amount or concentration of the biological component with high accuracy.

[0030] Note that the difference between the position CP(A) and the reference position RP may be used instead of the position CP(A), and the difference between the position CP(B) and the reference position RP may be used instead of the position CP(B).

[0031] Modification of Embodiment 1 Figure 4 is a diagram showing the configuration of a biological component analyzer according to a modification of Embodiment 1. The biological component analyzer according to the modification of Embodiment 1 differs from the non-invasive component analyzer according to Embodiment 1 in that the biological component analyzer according to the modification of Embodiment 1 includes an optical chopper 20 between the excitation light source 1 and the optical medium 3, and a lock-in amplifier 21 between the light position detector 4 and the calculator 11.

[0032] The optical chopper 20 is disposed in the optical path of the excitation light 6. The optical chopper 20 chops the excitation light 6 (continuous light) emitted from the excitation light source 1 at an arbitrary frequency. The excitation light 6 becomes intermittent light (pulsed light) that turns on and off periodically according to the chopping frequency (frequency at which light is turned on and off) of the optical chopper 20, and is incident on the optical medium 3. A well-known configuration can be applied to the optical chopper 20. The optical chopper 20 includes, for example, a rotating disk having openings that allow the excitation light 6 to pass and light-blocking portions that block the excitation light 6 arranged in the circumferential direction, and a motor that rotates the rotating disk. By periodically rotating the rotating disk using the motor, it is possible to switch whether or not the excitation light 6 is irradiated onto the sample 5. In other words, the excitation light 6 is intensity-modulated at the chopping frequency of the optical chopper 20. The chopping frequency of the excitation light 6 is determined by the rotation speed of the rotating disk.

[0033] The optical chopper 20 and the lock-in amplifier 21 are connected to an oscillator (not shown). The oscillator sets the chopping frequency (modulation frequency) of the optical chopper 20. The oscillator generates a control signal for chopping control of the pump light 6, and provides the generated control signal to the optical chopper 20 and the lock-in amplifier 21. The control signal includes the chopping frequency of the optical chopper 20.

[0034] The lock-in amplifier 21 selectively amplifies a signal synchronized with the chopping frequency (modulation frequency) of the optical chopper 20, among signals relating to the position of the probe light 7 output from the light position detector 4. The on-period of the chopping cycle corresponds to the period during which the excitation light 6 is irradiated. The off-period of the chopping cycle corresponds to the period during which the excitation light 6 is not irradiated.

[0035] Embodiment 2. Consider a case where the biological material to be measured is composed of multiple biological components, and it is desired to measure the amount or concentration of one of these biological components. For example, if the biological material is human skin, the excitation light 6 is absorbed by multiple components in the skin, such as sugar, oil, and sweat components, making it difficult to measure the amount or concentration of one biological component by measuring only the skin.

[0036] A reference substance is selected that absorbs the wavelength band of the excitation light source 1, has an absorption peak of one or more biological components in that wavelength band, and is in close contact with the optical medium 3. For example, sebum deposits on the surface of a living body or sweat itself can be used as the reference substance. By using the measurement results when the reference substance is sample 5, it is possible to eliminate the contribution of biological components other than the biological component to be measured. The refractive index gradient formed when the reference substance is sample 5 is fundamentally different from the refractive index gradient formed when the biological material is sample 5.

[0037] FIG. 5 is a flowchart showing the steps of a method for measuring a biological component using the biological component measuring device of the second embodiment.

[0038] In step S201, the sample 5 is set to a reference substance (e.g., water). In step S202, in a changing state in which the excitation light 6 from the excitation light source 1 is irradiated onto the sample 5, the light position detector 4 detects the position CP(A) of the emitted probe light 7b(A).

[0039] In step S203, sample 5 is set as a reference material (for example, sebum deposits or sweat on the surface of a living body).

[0040] In step S204, in a changing state in which the excitation light 6 from the excitation light source 1 is irradiated onto the sample 5, the light position detector 4 detects the position CP(C) of the emitted probe light 7b(C).

[0041] In step S205, the sample 5 is set on a biological material (for example, skin). In step S206, in a changing state in which the excitation light 6 from the excitation light source 1 is irradiated onto the sample 5, the light position detector 4 detects the position CP(B) of the emitted probe light 7b(B).

[0042] In step S207, the calculator 11 calculates the amount or concentration of a component of the biological material (for example, sugar in the skin) based on the positions CP(A), CP(B), and CP(C). For example, the calculator 11 calculates CP(B)-CP(A)-CP(C) to obtain the amount or concentration of a component of the biological material (for example, sugar in the skin).

[0043] In addition, instead of position CP(A), the difference between position CP(A) and the reference position RP may be used, instead of position CP(B), the difference between position CP(B) and the reference position RP may be used, and instead of position CP(C), the difference between position CP(C) and the reference position RP may be used.

[0044] Embodiment 3. In embodiment 1, it was shown that by performing a set of measurements using a reference substance as sample 5 and a set of measurements using a biological substance as sample 5, it is possible to measure the amount or concentration of a biological component in the biological substance after eliminating the influence of factors external to the excitation light source 1. In this embodiment, the detected CP(A) and CP(B) are used to more accurately calculate the amount or concentration of a component of the biological substance (for example, sugar in the skin).

[0045] 6 is a diagram showing the configuration of the computing unit 11. The computing unit 11 includes a learning device 91, an inference device 95, and a trained model storage device 94.

[0046] 7 is a diagram showing the configuration of a learning device 91 according to embodiment 3. The learning device 91 includes a data acquisition unit 92 and a model generation unit 93.

[0047] The data acquiring unit 92 acquires learning data including the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(B) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a biological substance, and the amount or concentration BA of the biological component. The learning data is data that associates the position CP(A) of the emitted probe light 7b(A) and the position CP(B) of the emitted probe light 7b(A) with the amount or concentration (correct answer) BA of the biological component.

[0048] The model generation unit 93 uses the training data to generate a trained model for inferring the amount or concentration BA of a biological component from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance.

[0049] The learning device 91 and the inference device 95 are not limited to those built into the computing unit 11. For example, the learning device 91 and the inference device 95 may be connected to the biological component measuring device via a network. The learning device 91 and the inference device 95 may also exist on a cloud server.

[0050] The learning algorithm used by the model generation unit 93 may be a known algorithm such as supervised learning, unsupervised learning, or reinforcement learning. As an example, a case where a neural network is applied will be described.

[0051] The model generation unit 93 performs so-called supervised learning, for example, according to a neural network model. Here, supervised learning refers to a technique in which a learning device is provided with pairs of input and result (label) data, and the learning device learns the features of the training data and infers the result from the input. A neural network is composed of an input layer consisting of multiple neurons, an intermediate layer (hidden layer) consisting of multiple neurons, and an output layer consisting of multiple neurons. The intermediate layer may have one or more layers. For example, in a three-layer neural network, when multiple inputs are input to the input layer, the input value is multiplied by a weight W1 and input to the intermediate layer, and the result is further multiplied by a weight W2 and output from the output layer. This output result varies depending on the values ​​of the weights W1 and W2. The neural network performs so-called supervised learning according to the training data acquired by the data acquisition unit 92. That is, the neural network learns by inputting into the input layer the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, and adjusting the weights W1 and W2 so that the result output from the output layer approaches the amount or concentration BA (correct answer) of the biological component.

[0052] The model generation unit 93 generates and outputs a trained model by performing the above-described learning.

[0053] The trained model storage device 94 stores the trained model output from the model generation unit 93.

[0054] 8 is a diagram showing the configuration of an inference device 95 according to embodiment 3. The inference device 95 includes a data acquisition unit 96 and an inference unit 97.

[0055] The data acquisition unit 96 acquires the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance.

[0056] The inference unit 97 uses a trained model stored in the trained model storage device 94 to infer the amount or concentration BA of the biological component from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, and infers the amount or concentration BA of the biological component from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, which are acquired by the data acquisition unit 96.

[0057] In this embodiment, the inference unit 97 uses a trained model trained by the model generation unit 93, but may also use a trained model generated outside the biological component measuring device.

[0058] In this embodiment, a case where supervised learning is applied to the learning algorithm used by the model generation unit 93 has been described, but the present invention is not limited to this. As for the learning algorithm, reinforcement learning, unsupervised learning, semi-supervised learning, or the like can also be applied in addition to supervised learning.

[0059] The learning algorithm used in the model generation unit 93 may be deep learning, which learns to extract the features themselves, or machine learning may be performed according to other known methods, such as genetic programming, functional logic programming, or support vector machines.

[0060] According to this embodiment, it is possible to measure the amount or concentration of a biological component to be measured with high accuracy.

[0061] Fourth Embodiment In this embodiment, the computing unit 11 includes a learning device 91A, an inference device 95A, and a trained model storage device 94A.

[0062] 9 is a diagram showing the configuration of a learning device 91A according to embodiment 4. The learning device 91A includes a data acquisition unit 92A and a model generation unit 93A.

[0063] The data acquisition unit 92A acquires learning data including the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, and the amount or concentration BA of the biological component.

[0064] The model generation unit 93A uses the training data to generate a trained model for inferring the amount or concentration BA of a biological component from the position CP(A) of the emitted probe light 7b(A) detected by the optical position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the optical position detector 4 when the sample 5 is a standard substance, and the position CP(B) of the emitted probe light 7b(B) detected by the optical position detector 4 when the sample 5 is a biological substance.

[0065] The trained model storage device 94A stores the trained model output from the model generation unit 93A.

[0066] 10 is a diagram showing the configuration of an inference device 95A according to embodiment 4. The inference device 95A includes a data acquisition unit 96A and an inference unit 97A.

[0067] The data acquisition unit 96A acquires the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance.

[0068] The inference unit 97A uses a trained model for inferring the amount or concentration BA of a biological component from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, which are acquired by the data acquisition unit 96A, to infer the amount or concentration BA of a biological component from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, which are acquired by the data acquisition unit 96A.

[0069] According to this embodiment, it is possible to measure the amount or concentration of a biological component to be measured with high accuracy.

[0070] Fifth Embodiment The computing unit 11 of the fifth embodiment includes a learning device 91B, an inference device 95B, and a trained model storage device 94B.

[0071] 11 is a diagram showing the configuration of a learning device 91B according to embodiment 5. The learning device 91B includes a data acquisition unit 92B, a preprocessing unit 98B, and a model generation unit 93B.

[0072] The data acquisition unit 92B acquires data including the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, and the amount or concentration BA of the biological component.

[0073] The preprocessing unit 98B calculates an index α from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference material, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological material. The index α is data that reveals biological components. The index α can be, for example, CP(B)-CP(A) or CP(B) / CP(A). The preprocessing unit 98B can reduce the amount of data and time required for learning, and can also avoid overlearning.

[0074] The model generation unit 93B uses learning data including the index α and the amount or concentration BA of the biological component to generate a trained model for inferring the amount or concentration BA of the biological component from the index α.

[0075] The trained model storage device 94B stores the trained model output from the model generation unit 93B.

[0076] 12 is a diagram showing the configuration of an inference device 95B according to embodiment 5. The inference device 95B includes a data acquisition unit 96B, a preprocessing unit 99B, and an inference unit 97B.

[0077] The data acquisition unit 96B acquires the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance.

[0078] The pre-processing unit 99B calculates an index α from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference material, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological material.

[0079] The inference unit 97B infers the amount or concentration BA of a biological component from the index α calculated by the preprocessing unit 99B using a trained model for inferring the amount or concentration BA of a biological component from the index α, which is stored in the trained model memory device 94B.

[0080] According to this embodiment, it is possible to reduce the amount of data and time required for learning, and also to easily measure the amount or concentration of a biological component to be measured with high accuracy.

[0081] Sixth Embodiment The computing unit 11 of the sixth embodiment includes a learning device 91C, an inference device 95C, and a trained model storage device 94C.

[0082] 13 is a diagram showing the configuration of a learning device 91C according to embodiment 6. The learning device 91C includes a data acquisition unit 92C, a preprocessing unit 98C, and a model generation unit 93C.

[0083] The data acquisition unit 92C acquires data including the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance, and the amount or concentration of the biological component.

[0084] The pre-processing unit 98C calculates one index β from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance.

[0085] The model generation unit 93C uses learning data including the index β and the amount or concentration BA of the biological constituent to generate a trained model for inferring the amount or concentration BA of the biological constituent from the index β.

[0086] The trained model storage device 94C stores the trained model output from the model generation unit 93C.

[0087] 14 is a diagram showing the configuration of an inference device 95C according to embodiment 6. The inference device 95C includes a data acquisition unit 96C, a preprocessing unit 99C, and an inference unit 97C.

[0088] The data acquisition unit 96C acquires the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference substance, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard substance, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological substance.

[0089] The pre-processing unit 99C calculates one index β from the position CP(A) of the emitted probe light 7b(A) detected by the light position detector 4 when the sample 5 is a reference material, the position CP(C) of the emitted probe light 7b(C) detected by the light position detector 4 when the sample 5 is a standard material, and the position CP(B) of the emitted probe light 7b(B) detected by the light position detector 4 when the sample 5 is a biological material. The index β can be, for example, CP(B)-CP(A)-CP(C).

[0090] The inference unit 97C infers the amount or concentration BA of a biological component from the index β calculated by the preprocessing unit 99C using a trained model for inferring the amount or concentration BA of a biological component from the index β stored in the trained model memory device 94C.

[0091] According to this embodiment, it is possible to reduce the amount of data and time required for learning, and also to easily measure the amount or concentration of a biological component to be measured with high accuracy.

[0092] In the first to sixth embodiments, the operation of the computing unit 11 may be implemented by hardware or software of a digital circuit.

[0093] 15 is a diagram showing a configuration in which the functions of the computing unit 11 are realized using software. The computing unit 11 includes a processor 1001 and a memory 1000 connected to a bus 1002. The processor 1001 executes a program stored in the memory 1000.

[0094] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The scope of the present disclosure is defined by the claims, not by the above description, and is intended to include all modifications within the meaning and scope of the claims.

[0095] 1 Excitation light source, 2 Probe light source, 3 Optical medium, 4 Optical position detector, 5 Sample, 6 Excitation light, 7, 7a, 7b Probe light, 8 Refractive index gradient region, 11 Computing unit, 20 Optical chopper, 21 Lock-in amplifier, 91, 91A, 91B, 91C Learning device, 92, 92A, 92B, 92C, 96, 96A, 96B, 96C Data acquisition unit, 93, 93A, 93B, 93C Model generation unit, 94, 94A, 94B, 94C Learned model storage device, 95, 95A, 95B, 95C Inference device, 97A, 97B, 97C Inference unit, 98B, 98C, 99B, 99C Preprocessing unit, 1000 Memory, 1001 Processor, 1002 Bus.

Claims

1. A biological component measuring device comprising: an optical medium including a sample mounting surface; an excitation light source that travels through the optical medium towards a sample placed on the sample mounting surface and emits excitation light having an excitation wavelength of the sample; a probe light source that emits probe light that travels through the optical medium; a light position detector that detects a position of the outgoing probe light emitted from the optical medium; and a calculator that calculates the amount or concentration of a biological component in the biological material based on the position of the outgoing probe light detected by the light position detector when the sample is a reference material, and the position of the outgoing probe light detected by the light position detector when the sample is a biological material, wherein the reference material absorbs the excitation light but does not have a specific absorption peak within the excitation wavelength of the excitation light.

2. The biological component measuring device of claim 1, wherein the computing unit is equipped with a learning device, the learning device including: a data acquiring unit that acquires learning data including the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the biological material, and the amount or concentration of the biological component; and a model generating unit that uses the learning data to generate a learned model for inferring the amount or concentration of the biological component from the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material.

3. The biological component measuring device of claim 1, wherein the computing unit is equipped with a learning device comprising: a data acquiring unit that acquires data including a position of the emitted probe light detected by the light position detector when the sample is the reference material, a position of the emitted probe light detected by the light position detector when the sample is the biological material, and an amount or concentration of the biological component; a pre-processing unit that calculates an index from the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material; and a model generating unit that uses learning data including the index and the amount or concentration of the biological component to generate a trained model for inferring the amount or concentration of the biological component from the index.

4. The biological component measuring device of claim 1, wherein the computing unit is equipped with an inference device, the inference device including: a data acquisition unit that acquires the position of the emitted probe light detected by the light position detector when the sample is the reference material and the position of the emitted probe light detected by the light position detector when the sample is the biological material; and an inference unit that infers the amount or concentration of the biological component from the position of the emitted probe light detected by the light position detector when the sample is the reference material and the position of the emitted probe light detected by the light position detector when the sample is the biological material, acquired by the data acquisition unit, using a trained model for inferring the amount or concentration of the biological component from the position of the emitted probe light detected by the light position detector when the sample is the reference material and the position of the emitted probe light detected by the light position detector when the sample is the biological material.

5. The biological component measuring device according to claim 1, wherein the computing unit is equipped with an inference device comprising: a data acquisition unit that acquires the position of the emitted probe light detected by the light position detector when the sample is the reference material and the position of the emitted probe light detected by the light position detector when the sample is the biological material; a pre-processing unit that calculates an index from the position of the emitted probe light detected by the light position detector when the sample is the reference material and the position of the emitted probe light detected by the light position detector when the sample is the biological material; and an inference unit that infers the amount or concentration of the biological component from the index calculated by the pre-processing unit using a trained model for inferring the amount or concentration of the biological component from the index.

6. The biological component measuring device of claim 1, wherein the calculator further calculates the amount or concentration of the biological component based on the position of the emitted probe light detected by the light position detector when the sample is a reference substance, the reference substance absorbing the excitation light and having one or more absorption peaks of the biological component within the excitation wavelength of the excitation light.

7. The biological component measuring device of claim 6, wherein the computing unit is equipped with a learning device, the learning device including: a data acquiring unit that acquires learning data including the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the biological material, and an amount or concentration of the biological component; and a model generating unit that uses the learning data to generate a learned model for inferring the amount or concentration of the biological component from the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material.

8. The biological component measuring device according to claim 6, wherein the computing unit is equipped with a learning device, the learning device including: a data acquiring unit that acquires data including a position of the emitted probe light detected by the light position detector when the sample is the reference material, a position of the emitted probe light detected by the light position detector when the sample is the reference material, a position of the emitted probe light detected by the light position detector when the sample is the biological material, and an amount or concentration of the biological component; a pre-processing unit that calculates an index from the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material; and a model generating unit that uses learning data including the index and the amount or concentration of the biological component to generate a trained model for inferring the amount or concentration of the biological component from the index.

9. The computing unit includes an inference device, the inference device including: a data acquisition unit that acquires a position of the emitted probe light detected by the light position detector when the sample is the reference material, a position of the emitted probe light detected by the light position detector when the sample is the standard material, and a position of the emitted probe light detected by the light position detector when the sample is the biological material; and an inference unit that infers the amount or concentration of the biological component from the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the biological material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material, obtained by the data acquisition unit, using a trained model for inferring the amount or concentration of the biological component from the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material.

10. The computing unit is equipped with an inference device, the inference device including: a data acquisition unit that acquires the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material; a pre-processing unit that calculates an index from the position of the emitted probe light detected by the light position detector when the sample is the reference material, the position of the emitted probe light detected by the light position detector when the sample is the reference material, and the position of the emitted probe light detected by the light position detector when the sample is the biological material; and an inference unit that infers the amount or concentration of the biological component from the index calculated by the pre-processing unit using a trained model for inferring the amount or concentration of the biological component from the index.

11. A biological component measuring device as described in any one of claims 1 to 10, wherein, in a planar view of the sample placement surface, the optical path of the probe light in the optical medium overlaps with a portion of the sample placement surface that is irradiated by the excitation light.

12. A method for measuring a biological component comprising the steps of: an excitation light source emitting excitation light having an excitation wavelength of the sample, traveling through the optical medium toward a sample placed on a sample placement surface of the optical medium; a probe light source emitting probe light traveling through the optical medium; a light position detector detecting a position of an outgoing probe light of the probe light incident on the optical medium when the sample is a reference material; a light position detector detecting a position of an outgoing probe light of the probe light incident on the optical medium when the sample is a biological material; and a calculator calculating an amount or concentration of a biological component in the biological material based on the position of the outgoing probe light detected by the light position detector when the sample is the reference material and the position of the outgoing probe light detected by the light position detector when the sample is the biological material, wherein the reference material absorbs the excitation light and does not have a specific absorption peak within the excitation wavelength of the excitation light.

13. A method for measuring a biological component as described in claim 12, further comprising a step in which the light position detector detects a position of an outgoing probe light of the probe light incident on the optical medium when the sample is a reference substance, and the step of calculating the amount or concentration of the biological component includes a step in which the calculator calculates the amount or concentration of the biological component based on the position of the outgoing probe light detected by the light position detector when the sample is the reference substance, the position of the outgoing probe light detected by the light position detector when the sample is the reference substance, and the position of the outgoing probe light detected by the light position detector when the sample is the biological substance, and wherein the reference substance absorbs the excitation light and has absorption peaks of one or more biological components within the excitation wavelength of the excitation light.

Images