Information processing device, information processing method, control program, and recording medium

Abstract

To provide an information processing device capable of automatically evaluating surface characteristics of a sample.SOLUTION: An information processing device (1) includes: an acquisition section (11) that acquires a captured image obtained by capturing a sample and a liquid when a gas is jetted to the liquid covering the sample; a formation region prediction section (12) that inputs the captured image to a prediction model (22) to predict information on a formation region formed by movement of at least part of the liquid covering the sample due to the jet of the gas; and an output control section (13) that causes an output device (3) to output a prediction result.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information processing apparatus and an information processing method for analyzing the characteristics of a sample surface. [Background technology]

[0002] Commonly used methods for evaluating solid surfaces include surface analysis methods such as time-of-flight secondary ion mass spectrometry (TOF-SIMS), X-ray photoelectron spectroscopy (XPS), infrared spectroscopy (IR), Raman spectroscopy, and contact angle methods. However, when evaluating the surfaces of medical devices, etc., it is difficult to evaluate the surfaces in the liquid environment in which the medical devices, etc. are actually used, and it is difficult to achieve an appropriate evaluation.

[0003] As a method for evaluating a solid surface in a liquid environment, a method has been disclosed in which a gas is sprayed onto a substance in liquid to remove the liquid and evaluate the wettability of the substance (see, for example, Patent Document 1, etc.). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. 2013 / 176264 Summary of the Invention [Problem to be solved by the invention]

[0005] However, while the conventional techniques described above can evaluate the surfaces of planar solid samples on the order of a few centimeters in size, such as cell sheets and culture dishes, they cannot properly evaluate the characteristics of sample surfaces with narrow curved surfaces, such as guidewires and catheters, in a liquid environment.

[0006] Furthermore, in the conventional techniques, the analysis and evaluation of the sample surface had to be performed manually, leaving room for improvement in terms of work costs and the types of sample surfaces that could be analyzed.

[0007] According to one aspect of the present invention, an information processing device, an information processing method, and the like that can automatically evaluate the characteristics of a sample surface can be provided. [Means for solving the problem]

[0008] In order to solve the above problems, an information processing device according to one aspect of the present invention includes an acquisition unit that acquires an image of a specimen and a liquid when gas is injected into a liquid covering the specimen, a formation region prediction unit that inputs the image into a prediction model and predicts information about a formation region formed by the movement of at least a portion of the liquid covering the specimen by the injection of the gas, and an output control unit that outputs the prediction result by the formation region prediction unit to an output device, and the prediction model is trained by machine learning using teacher data including an explanatory variable including a specimen image of the specimen specimen and the liquid when gas is injected into the liquid covering the specimen specimen, and a target variable including information about a specimen region that is identified in advance by a person having a predetermined skill based on the specimen image and is formed by the movement of at least a portion of the liquid covering the specimen specimen.

[0009] In addition, an information processing device according to one aspect of the present invention includes an acquisition unit that acquires an image of the specimen and the liquid when gas is injected onto the liquid covering the specimen, a first formation region prediction unit that inputs the image into a first prediction model to predict a formation region formed by the movement of at least a portion of the liquid covering the specimen by the injection of the gas, a second formation region prediction unit that inputs region information indicating the predicted formation region into a second prediction model to predict information regarding the formation region, and an output control unit that outputs a prediction result by the second formation region prediction unit to an output device, wherein the first prediction model is trained by machine learning using teacher data including an explanatory variable including a specimen image of the specimen and the liquid when gas is injected onto the liquid covering the specimen sample, and an explanatory variable including region information indicating a specimen region identified in advance based on the specimen image by a person having a predetermined skill, the specimen region being formed by the movement of at least a portion of the liquid covering the specimen sample, and the second prediction model is trained by machine learning using teacher data including an objective variable including the region information, and an objective variable including information regarding the specimen region determined in advance by a person having a predetermined skill.

[0010] An information processing method according to one aspect of the present invention includes an acquisition step of acquiring an image of a specimen and a liquid when gas is injected into a liquid covering the specimen, a formation region prediction step of inputting the acquired image into a prediction model to predict information about a formation region formed by the movement of at least a portion of the liquid covering the specimen by the injection of the gas, and an output control step of causing an output device to output a prediction result by the formation region prediction unit, wherein the prediction model is trained by machine learning using teacher data including an explanatory variable including a specimen image of the specimen specimen and the liquid when gas is injected into the liquid covering the specimen specimen, and a target variable including information about a specimen region identified in advance by a person having a predetermined skill based on the specimen image, the specimen region being formed by the movement of at least a portion of the liquid covering the specimen specimen.

[0011] Moreover, an information processing method according to one aspect of the present invention includes an acquisition step of acquiring an image of a sample and a liquid when gas is ejected onto the liquid covering the sample, a first formation region prediction step of inputting the captured image into a first prediction model and predicting a formation region formed by movement of at least a part of the liquid covering the sample by the ejection of the gas, a second formation region prediction step of inputting region information indicating the predicted formation region into a second prediction model and predicting information regarding the formation region, and an output control step of causing an output device to output a prediction result by the second formation region prediction unit, is trained by machine learning using training data including explanatory variables including a specimen image of the specimen and the liquid when gas is injected into the liquid covering the specimen, and a target variable including area information indicating a specimen area identified in advance based on the specimen image by a person having specified skills, the specimen area being formed by the movement of at least a portion of the liquid covering the specimen, and the second prediction model is trained by machine learning using training data including explanatory variables including the area information, and a target variable including information about the specimen area determined in advance by a person having specified skills.

[0012] In addition, the information processing device according to each aspect of the present disclosure may be realized by a computer. In this case, the control program of the information processing system that realizes the information processing system on a computer by causing the computer to operate as each part (software element) of the information processing system, and the computer-readable recording medium on which it is recorded, also fall within the scope of the present disclosure. Effect of the Invention

[0013] According to one aspect of the present invention, the characteristics of the surface of a sample can be automatically evaluated. [Brief description of the drawings]

[0014] [Figure 1]1 is a functional block diagram showing an example of a configuration of an information processing device according to an embodiment of the present invention; [Diagram 2] 1A and 1B are diagrams illustrating a schematic view of gas being ejected onto a surface of a sample covered with a liquid. [Diagram 3] FIG. 13 is a diagram showing an example of a captured image of a formation region. [Figure 4] FIG. 2 is a cross-sectional view showing a schematic view of a formation region. [Diagram 5] FIG. 13 is a schematic diagram for explaining the length of a formation region. [Figure 6] 11 is a flowchart showing an example of a flow of a process executed by an information processing device. [Figure 7] FIG. 11 is a functional block diagram showing an example of a configuration of an information processing device according to another embodiment of the present invention. [Figure 8] 11 is a flowchart showing an example of a flow of a process executed by an information processing device. [Figure 9] FIG. 11 is a functional block diagram showing an example of a configuration of an information processing device according to another embodiment of the present invention. [Figure 10] 11 is a flowchart showing an example of a flow of a process executed by an information processing device. [Figure 11] 1 is a graph showing measurement results in an example of the present invention. [Figure 12] 1 is a graph showing measurement results in an example of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] The following describes in detail the embodiments of the present invention. However, the present invention is not limited to this, and various modifications are possible within the scope described, and embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. All academic documents and patent documents described in this specification are incorporated herein by reference. Furthermore, unless otherwise specified in this specification, "A to B" representing a numerical range means "A or more (including A and greater than A) and B or less (including B and smaller than B)."

[0016] [Embodiment 1] An information processing device according to an embodiment of the present invention is a device for evaluating surface characteristics of a sample. In the following, an information processing device communicably connected to one imaging device and one output device will be described as an example. However, the information processing device is not limited to this configuration. For example, the number of imaging devices and output devices connected to the information processing device may be more than one. In addition, the information processing device may not be connected to at least one of the imaging device and the output device.

[0017] (Configuration of information processing device 1) Hereinafter, a configuration of an information processing device 1 according to an embodiment of the present invention will be described with reference to Fig. 1. Fig. 1 is a functional block diagram showing an example of the configuration of the information processing device 1.

[0018] 1 includes a storage unit 20 that stores various data used by the information processing device, and a control unit 10 that controls each unit of the information processing device, but is not limited to this configuration. For example, the storage unit 20 may be an external device attached to the information processing device 1.

[0019] <Storage section 20> First, a description will be given of the storage unit 20. In the storage unit 20, image data 21 and a prediction model 22 may be stored.

[0020] The image data 21 is a captured image of the sample and the liquid when gas is sprayed onto the liquid covering the sample. The captured image may be an image captured by the imaging device 2.

[0021] The prediction model 22 is trained by machine learning using training data. The training data includes the following explanatory variables and objective variables. The explanatory variables include images of the specimen and the liquid when gas is sprayed onto the liquid covering the specimen. The objective variable includes information about a specimen region that is identified in advance by a person with a predetermined skill based on a specimen image and is formed by the movement of at least a part of the liquid covering the specimen sample. Note that the type of information about the specimen region may be the same as the information about the formed region described above.

[0022] A method of modeling the relationship between such explanatory variables and objective variables includes, for example, a method using a statistical model. The prediction model 22 may be a linear model or a nonlinear model. For example, the prediction model 22 may be constructed by the least squares method, principal component regression, neural network, k-NN (k-nearest neighbor algorithm), SVR (Support Vector Regression), or the like.

[0023] <Control unit 10> Next, the control unit 10 will be described. The control unit 10 includes an acquisition unit 11, a formation region prediction unit 12, and an output control unit 13. In addition, some of the blocks included in the control unit 10 may be omitted from the control unit 10 by giving the functions of the blocks to another device capable of communicating with the information processing device 1. For example, the function of the output control unit 13 may be given to the other device. In this case, the information processing device 1 may output the result predicted by the formation region prediction unit 12 through the other device.

[0024] The acquisition unit 11 acquires an image of the sample and the liquid when gas is sprayed onto the liquid covering the sample. The acquisition unit 11 may be configured to acquire images acquired by the imaging device 2 in real time, or may be configured to read out the captured images from image data 21 stored in the storage unit 10.

[0025] Here, the imaging device 2 is a device capable of imaging the sample and liquid when gas is injected into the liquid covering the sample. The imaging device 2 may be, for example, a digital camera, a digital video, a mobile terminal with an imaging function (for example, a smartphone, a tablet, a laptop, a mobile phone, etc.), a device that injects gas into the liquid covering the sample, or an imaging device additionally provided with the device that injects gas. From the viewpoint of being able to simultaneously inject gas and image capture, it is preferable that the imaging device 2 is a device that injects gas into the liquid covering the sample.

[0026] The formation region prediction unit 12 inputs the image acquired by the acquisition unit 11 into the prediction model 22 to predict information about the formation region. The formation region prediction unit 12 may be configured to predict at least one of the dimensions, area, shape, and position of the formation region as information about the formation region. More specifically, the information about the formation region may be numerical information such as the x-coordinate and y-coordinate of the formation region, or visual information showing an area corresponding to the formation region in the captured image on the captured image. The shape of the formation region illustrated in the information about the formation region is not particularly limited. Examples of the shape of such a formation region include any shape such as a line segment, a circle, an ellipse, and a rectangle.

[0027] The output control unit 13 outputs the prediction result predicted by the formation region prediction unit 12 to the output device 3. The output mode of the output device 3 is not particularly limited, and may be, for example, a display output, a print output, or an audio output.

[0028] The captured image acquired by the acquisition unit 11 of the information processing device according to one embodiment of the present invention is an image of the sample and liquid when gas is injected into the liquid covering the sample. Here, the injection of gas into the liquid covering the sample can be expressed in other words as the gas being injected from outside the liquid onto the surface of the sample covered by the liquid. Here, the state in which the surface of the sample is covered by the liquid means that the part of the surface of the sample that is to be imaged is covered by the liquid, and the part that is not to be imaged may or may not be covered by the liquid. For example, the gas may be injected into the liquid covering the sample while the sample is immersed in the liquid.

[0029] The surface of the sample may be covered with liquid, and the amount of liquid covering the surface is not particularly limited, but the thickness of the liquid layer covering the sample is, for example, 1 μm to 10.0 mm, more preferably 0.1 mm to 5.0 mm, and even more preferably 1.0 mm to 3.0 mm. If the thickness of the liquid layer is 1 μm or more, it is a sufficient thickness for evaluating the characteristics of the sample covered with the liquid. In addition, if the thickness of the liquid layer is 10.0 mm or less, it is preferable because at least a part of the liquid present on the surface of the sample can be moved by gas injection.

[0030] The liquid is not particularly limited, and for example, when wettability is evaluated, a liquid for evaluating wettability may be used, and when the adsorption behavior of a solute is evaluated, a liquid containing the solute for evaluating the adsorption behavior may be used.

[0031] The liquid may be an aqueous liquid such as water, an aqueous solution, or an aqueous dispersion. The water may be tap water; pure water such as desalted water, RO water, or distilled water; or ultrapure water. The aqueous solution and the aqueous dispersion may each contain a solvent other than water and a dispersing medium other than water. The solvent other than water and the dispersing medium other than water may be an organic solvent or a hydrocarbon containing a functional group such as a hydroxyl group, an aldehyde group, a carbonyl group, a carboxyl group, a nitro group, an amino group, a sulfo group, an ether group, an ester group, or a halogen. More specific examples of the solvent other than water and the dispersing medium other than water may include, but are not limited to, alcohol such as ethanol, acetone, ethyl acetate, ethylene glycol, tetrahydrofuran (THF), dibutyl ether, dimethyl sulfoxide (DMSO), dimethylformamide (DMF), acetonitrile, aniline, nitrobenzene, cyclohexane, isooctane, 1,4-dioxane, N-methylformamide, N-methylpyrrolidone (NMP), and dichloromethane. These solvents other than water may be used alone or in combination of two or more kinds. The aqueous solution and the aqueous dispersion may contain solutes such as surfactants, salts, particles, proteins, amino acids, peptides, micelles, emulsions, cells, viruses, bacteria, DNA, mRNA, microorganisms, low molecular weight organic matter, polymers, inorganic acids, and inorganic bases. Examples of the surfactants include anionic surfactants, cationic surfactants, amphoteric surfactants, and nonionic surfactants. Examples of the salts include sodium chloride and potassium chloride. Examples of the particles include carbon, polymers, silica, alumina, and titanium oxide. The surfactants, salts, and particles may be used alone or in combination of two or more kinds. Alternatively, the liquid may be an organic solvent. In such a case, the organic solvent may be, but is not limited to, an organic solvent or hydrocarbon containing a functional group such as a hydroxyl group, an aldehyde group, a carbonyl group, a carboxyl group, a nitro group, an amino group, a sulfo group, an ether group, an ester group, or a halogen.More specific examples of the organic solvent include alcohols such as ethanol, acetone, ethyl acetate, ethylene glycol, tetrahydrofuran (THF), dibutyl ether, dimethyl sulfoxide (DMSO), dimethylformamide (DMF), acetonitrile, aniline, nitrobenzene, cyclohexane, isooctane, 1,4-dioxane, N-methylformamide, N-methylpyrrolidone (NMP), dichloromethane, etc. These organic solvents may be used alone or in combination of two or more kinds.

[0032] The liquid may be, for example, a liquid present in the human or animal body, such as blood, plasma, lymph, tissue fluid, etc., or a simulated liquid (or substitute liquid), such as physiological saline, drip infusion, infusion, artificial blood, liquid culture medium, etc. For example, since medical devices are used inside the human or animal body, the properties of the surface can be evaluated when the surface is covered with such a liquid.

[0033] Another example of the liquid is a liquid containing a solute. By using such a liquid, the adsorption behavior of the solute on the surface of the sample can be evaluated. The solute is not particularly limited, and any substance whose adsorption behavior on the surface is to be evaluated may be used. Examples of the solute include amino acids; peptides such as daptomycin and lysozyme; proteins such as bovine serum albumin (BSA) and collagen; surfactants such as anionic surfactants, cationic surfactants, amphoteric surfactants, and nonionic surfactants such as Tween; and phospholipids such as phosphatidylcholine. The solute may be used alone or in combination of two or more kinds. The solvent for dissolving the solute is also not particularly limited, and water, a combination of water and the above-mentioned solvent other than water, or the above-mentioned organic solvent may be used.

[0034] By injecting gas into the liquid covering the sample, at least a part of the liquid covering the sample is displaced. That is, at least a part of the liquid present on the surface of the sample is displaced by the injected gas. As a result of the movement of at least a part of the liquid, the thickness of the liquid layer present on the surface of the sample becomes thinner, or the liquid present on the surface of the sample is removed to expose the sample surface.

[0035] The sample may be a solid including a ridge. Here, the ridge may be a linear portion that protrudes in a ridge shape, and the ridge-like protrusion may be sharp or rounded. When the sample has a ridge, the gas is sprayed along the ridge in this step. As a result, at least a part of the liquid present on the surface of the sample, i.e., the liquid covering the ridge, which is the surface of the sample to be evaluated, moves linearly along the ridge. As a result of at least a part of the liquid moving linearly along the ridge, the layer of the liquid present on the surface of the sample becomes linearly thinner along the ridge, or the liquid present on the surface of the sample is removed linearly along the ridge to expose the surface of the sample. There is no limitation on the mechanical properties of the solid, such as hardness and viscoelasticity, and gels and sponges are also included.

[0036] The shape of the solid including the ridgeline portion is not particularly limited as long as it includes the ridgeline portion, and may be, for example, a columnar body including the ridgeline portion. Examples of such columnar bodies include, for example, a circular cylinder; an elliptical cylinder; a polygonal prism such as a hexagonal prism, a pentagonal prism, a square prism, or a triangular prism; and a columnar body having a shape with a protrusion on the bottom surface. In a cylinder or an elliptical cylinder, any line perpendicular to the bottom surface on the side surface is the ridgeline portion. In a polygonal prism, any line among the lines forming the boundary between two faces forming an angle on the side surface is the ridgeline portion. The columnar body may be a hollow tubular body. That is, the shape of the solid including the ridgeline portion may be a cylinder; an elliptical cylinder; or a cylindrical body of a polygonal prism such as a hexagonal prism, a pentagonal prism, a square prism, or a triangular prism. However, the shape of the solid including the ridgeline portion is not limited to the above-mentioned shapes, and may be a braided structure such as a braid or a rope; a chain shape, or the like. In addition, the ridgeline portion may include a structure in which fibers such as nonwoven fabric or filter paper are randomly entangled. Furthermore, the sample may be a porous solid so long as it is a solid that includes a ridge line portion. The porous solid may have closed pores or through holes inside. Alternatively, the shape of the solid that includes the ridge line portion may be a shape in which a sample having a dome-like shape is deformed to generate the ridge line portion. Here, the dome shape is not limited to a perfect semicircle, but may be a shape that has a cross-sectional arc shape. For example, medical devices such as catheters, guide wires, stents, injection needles, and hollow fibers have a cylindrical or columnar shape and can be suitably evaluated by the method according to one embodiment of the present invention. In addition, for example, contact lenses, more preferably soft contact lenses, have a dome-like shape and can be formed into a shape in which a ridge line portion is generated by applying a force in a certain direction when fixing the sample, and therefore can be evaluated by the method according to one embodiment of the present invention.

[0037] The dimensions of the sample are not particularly limited. For example, when the sample is columnar or cylindrical, the outer diameter is, for example, 0.01 mm to 10.0 mm, and may be 0.1 mm to 2.0 mm. If the outer diameter is 0.01 mm or more, the size is sufficient for evaluating the surface characteristics. Also, if the outer diameter is 10.0 mm or less, it is preferable because at least a part of the liquid present on the surface of the sample can be moved by gas injection.

[0038] The material of the sample is not particularly limited and may be organic, inorganic, or a mixture thereof, and may be, for example, plastic, glass, metal, rubber, wood, paper, ceramics, leather, or concrete. The sample may have a surface coating. The type of coating is not particularly limited and may be, for example, a hydrophilic coating or a water-repellent coating. Alternatively, the sample may have a surface treatment such as plating, chemical treatment, spraying such as plasma treatment, surface hardening, or the like. By evaluating the characteristics of the surface of a sample that has been surface-treated such as a coating, the information processing device according to one embodiment of the present invention may be used for the development of a coating film or the like.

[0039] Examples of the sample include medical devices and materials such as catheters, guide wires, stents, endoscopes, injection containers, injection needles, drip tubes, hollow fibers, and contact lenses; glass rods, straws, copper wires, metal tubes, vinyl ties, wiring cords, and syringes.

[0040] The sample is covered with the liquid and is placed so as not to move due to the gas to be sprayed, and the method of placement is not particularly limited. Examples of the method of placement include a method of immersing the sample in the liquid. In such a case, the sample may be placed by placing it on the bottom of a container containing the liquid, or the sample may be held in the liquid by a holding member. When the sample is placed on the bottom of a container containing the liquid, the sample does not need to be fixed as long as it does not move due to the gas to be sprayed. However, when the sample is placed on the bottom of a container containing the liquid, it is more preferable that the sample is fixed regardless of whether it moves due to the gas to be sprayed. The method of fixing the sample is also not particularly limited, and examples of the method include a method of fixing a portion of the sample other than the surface to be evaluated for characteristics, using a fixing member such as tape.

[0041] The sample may be one sample or multiple samples. By installing multiple samples and measuring the average value of each sample, average data of the samples can be obtained.

[0042] The gas is not particularly limited, but examples thereof include air, nitrogen, inert gases such as helium and argon, carbon dioxide, oxygen, etc. The gas may contain water vapor. In addition, two or more types of gas may be mixed and used. The gas may be cooled or heated before injection within a range in which the gas state is maintained.

[0043] The amount of gas injected may be an amount that displaces at least a part of the liquid covering the sample by the injected gas, and may be appropriately selected according to the material of the sample, the shape of the sample, the thickness of the liquid layer covering the surface of the sample, the type of liquid, the distance between the sample surface and the liquid layer surface and the gas injection port, the shape and size of the injection port, the type of gas, the temperature, and the like. The amount of gas injected is, for example, preferably 0.1 kPa to 20.0 kPa, more preferably 2.0 kPa to 10.0 kPa, and even more preferably 3.0 kPa to 7.0 kPa, such that the pressure applied to the surface of the sample is 0.1 kPa to 20.0 kPa, more preferably 2.0 kPa to 10.0 kPa, and even more preferably 3.0 kPa to 7.0 kPa. By setting the amount of gas injected within the above range, at least a part of the liquid covering the sample is displaced and moved, and the area formed by the movement of at least a part of the liquid covering the sample can be identified and its size can be measured, which is preferable. In addition, the pressure may be changed stepwise during injection for evaluation.

[0044] The gas may be sprayed from vertically above the surface of the sample covered with liquid, or from obliquely above, so long as the sprayed gas can displace at least a portion of the liquid covering the sample.

[0045] The shape of the gas jet is not particularly limited as long as the jetted gas can displace at least a portion of the liquid covering the sample, and the cross-sectional shape in a plane perpendicular to the jet direction can be circular, elliptical, linear, star-shaped, etc.

[0046] The time for which the gas is injected is not particularly limited and may be set appropriately depending on the property to be evaluated. The time for which the gas is injected is, for example, 0.5 seconds to 60 seconds, may be 1 second to 10 seconds, or may be 2 seconds to 4 seconds.

[0047] The method of injecting the gas is not particularly limited. For example, the gas can be injected using a nozzle. In this case, the inner diameter of the nozzle is not particularly limited and may be appropriately set depending on the characteristic to be evaluated. The inner diameter of the nozzle is, for example, 0.01 mm to 10.0 mm, and may be 0.01 mm to 0.5 mm. The distance between the liquid layer surface and the gas injection port is, for example, 1.0 mm to 20.0 mm.

[0048] 2 to 4 are diagrams showing an example of gas injection in this process and the movement of liquid by gas injection. In the example of FIG. 2, the sample R is immersed in the liquid L while being fixed to the bottom of the petri dish containing the liquid L by a fixing member. Then, the gas G is injected onto the surface of the sample 1 covered with the liquid L using a nozzle N. FIG. 3 is a photograph of a formation region P formed by linear movement of at least a part of the liquid covering the sample by gas injection. Also, FIG. 4 is a diagram showing a cross section of the moved liquid. The diagram showing the cross section of the moved liquid is a diagram showing the petri dish to which the sample R is fixed, viewed from the side of the petri dish so that the cross section of the sample R is the front. As shown in the figure, the sample R is cylindrical, and the side surface of the cylinder on the side where the gas G is injected is the above-mentioned ridge line portion. By injecting the gas G, the liquid present on the surface of the sample 1, i.e., at least a part of the liquid covering the ridge line portion, which is the surface to be evaluated of the sample R, moves linearly along the ridge line portion. 4, at least a part of the liquid L moves linearly along the ridge line, and as a result, as can be seen from the liquid level Q of the liquid present on the surface of the sample R, the liquid L present on the surface of the sample R is removed along the ridge line, exposing the sample surface. The removal of the liquid L along the ridge line is shown by a photograph of a formation region P that is formed by the movement of at least a part of the liquid covering the sample by the gas jet.

[0049] Fig. 5 is a schematic diagram showing a linear region formed by linear movement of at least a part of the liquid covering the sample by gas injection in a method according to one embodiment of the present invention. In the example shown in Fig. 5, when gas is injected onto a cylindrical or columnar sample, at least a part of the liquid covering the cylindrical or columnar sample, i.e., the liquid covering the ridgeline portion, which is the surface of the sample to be measured, moves linearly along the ridgeline portion to form a linear region. As time passes, the linear region extends longer in the longitudinal direction.

[0050] (Processing performed by information processing device 1) An overview of an information processing method according to one embodiment of the present invention will be described with reference to Fig. 6. Fig. 6 is a flowchart showing an overview of the information processing method performed by the information processing device 1.

[0051] In the acquisition step S1, an image of the sample and the liquid is acquired when the acquisition unit 11 injects gas into the liquid covering the sample. At this time, the image to be acquired may be an image acquired by the imaging device 30, or may be image data 21 stored in the storage unit 10.

[0052] Next, in the formation region prediction step S2, the formation region prediction unit 12 inputs the captured image to the prediction model 22 to predict information about the formation region formed by the movement of at least a part of the liquid covering the sample by the gas injection. At this time, the prediction model 22 is trained by machine learning using the training data. The training data includes. Details of the explanatory variables and the objective variables are as described above.

[0053] In the output control step S3, the output control unit 13 causes the output device 3 to output the result predicted by the formation region prediction unit 12.

[0054] Conventionally, obtaining information about the formation area in a captured image has been a time-consuming and cumbersome process, but by using the information processing device 1, it is possible to automatically predict information about the formation area based on the captured image.

[0055] [Embodiment 2] Hereinafter, the configuration of an information processing device 1a according to another embodiment of the present invention will be described with reference to FIG. 7. FIG. 7 is a block diagram showing an example of a main configuration of an information processing device 1a different from that of the first embodiment. The information processing device 1a includes a control unit 10a that controls each unit of the information processing device, and a storage unit 20a that stores various data used by the information processing device. The control unit 10 includes an acquisition unit 11, a first formation region prediction unit 12a, a second formation region prediction unit 12b, and an output control unit 13. The storage unit 20 stores image data 21, a first prediction model 22a, and a second prediction model 22b. Note that the description of the matters already described in the first embodiment will be omitted.

[0056] <Storage section 20a> First, the storage unit 20a will be described. The storage unit 20a may store image data 21, a first prediction model 22a, and a second prediction model 22b.

[0057] The first prediction model 22a is trained by machine learning using training data. The training data includes the following explanatory variables and objective variables. The explanatory variables include images of the specimen and the liquid when gas is sprayed onto the liquid covering the specimen. The objective variable includes area information indicating a specimen area that has been identified in advance by a person having a predetermined skill based on a specimen image and that has been formed by the movement of at least a portion of the liquid covering the specimen sample.

[0058] The region information predicted by the first prediction model 22a may be binarized as necessary. The first prediction model 22a may be a model that predicts the segmentation of an image, and specifically may be, for example, a Unet model.

[0059] The second prediction model 22b is trained by machine learning using training data. The training data includes the following explanatory variables and objective variables. The training data includes the following explanatory variables and objective variables. The explanatory variable is area information that indicates a specimen area formed by the movement of at least a part of the liquid covering the specimen sample. The objective variable includes information about the sample area that has been determined in advance by a person having a given skill.

[0060] The second prediction model 22b may be a model that performs image classification or regression, and specifically may be, for example, Resnet.

[0061] The first prediction model 22a and the second prediction model 22b may be modeled in a similar manner to the prediction model 22 described above.

[0062] <Control unit 10a> Next, the control unit 10a will be described. The control unit 10a includes an acquisition unit 11, a first formation region prediction unit 12a, a second formation region prediction unit 12b, and an output control unit 13.

[0063] The first formation region prediction unit 12a inputs the image acquired by the acquisition unit 11 into the first prediction model 22a and predicts the formation region formed by the movement of at least a part of the liquid covering the sample by the gas injection. In other words, the first formation region prediction unit 12a does not need to predict information about the formation region, and may only predict the range in the image that corresponds to the formation region.

[0064] The second formation area prediction unit 12b inputs area information indicating the formation area predicted by the first formation area 12a into the second prediction model 22b to predict information about the formation area. The information about the formation area predicted by the second formation area prediction unit is as described above.

[0065] The results of prediction by the first formation region prediction unit 12a and the second formation region prediction unit 12b may be output by the output control unit 13 to an output device.

[0066] (Processing performed by information processing device 1a) An overview of an information processing method according to one embodiment of the present invention will be described with reference to Fig. 8. Fig. 8 is a flowchart showing an overview of an information processing method performed by the information processing device 1a.

[0067] Step S1 is as described above.

[0068] In the first formation region prediction step S2a, the first formation region prediction unit 12a inputs the captured image into the first prediction model 22a to predict a formation region formed by the movement of at least a part of the liquid covering the sample by the gas injection. At this time, the prediction model 22a is trained by machine learning using training data. The training data includes explanatory variables and objective variables. The details of the explanatory variables and objective variables are as described above.

[0069] In the second formation region prediction step S2b, the second formation region prediction unit 12b inputs region information indicating the formation region predicted in step S2a into the second prediction model 22b to predict information about the formation region. At this time, the prediction model 22b is trained by machine learning using teacher data. The teacher data includes explanatory variables and objective variables. Details of the explanatory variables and objective variables are as described above.

[0070] In the output control step S3, the output control unit 13 causes the output device 3 to output the result predicted by the second formation region prediction unit 12b. Note that in the output control step S3, the output control unit 13 may also cause the output device 3 to output the result predicted by the first formation region prediction unit 12a.

[0071] By using the information processing device 1a, it is possible to automatically predict information on the formation region based on the captured image. In addition, by using the information processing device 1a, it is also possible to confirm the result predicted by the first formation region prediction unit 12a.

[0072] [Embodiment 3] (Configuration of information processing device 1b) Hereinafter, the configuration of an information processing device 1b according to another embodiment of the present invention will be described with reference to FIG. 9. FIG. 9 is a block diagram showing an example of a main configuration of an information processing device 1b different from the first and second embodiments. The information processing device 1b includes a control unit 10b that controls each unit of the information processing device, and a storage unit 20b that stores various data used by the information processing device. The control unit 10a includes an acquisition unit 11, a formation region prediction unit 12c, an output control unit 13b, and a prediction result correction unit 14. The storage unit 20 stores image data 21, a prediction model 22, and a prediction / correction result 23. Note that the description of the matters already described in the first or second embodiment will be omitted.

[0073] <Control unit 10b> The control unit 10b includes an acquisition unit 11, a formation region prediction unit 12c, an output control unit 13b, and a prediction result correction unit .

[0074] The formation area prediction unit 12c inputs the image acquired by the acquisition unit 11 into the prediction model 22 to predict information about the formation area. The formation area prediction unit 12c transmits the predicted information about the formation area not only to the output control unit 13b but also to the prediction result correction unit 14. The formation area prediction unit 12c may be replaced with the first formation area prediction unit 12a and the second formation area prediction unit 12b. In that case, the results predicted by the first formation area prediction unit 12a and the second formation area prediction unit 12b are transmitted to the output control unit 13b and the prediction result correction unit 14, respectively.

[0075] The prediction result correction unit 14 receives an input of a correction instruction for the prediction result by the formation region prediction unit 12c outputted to the output device 3. The prediction result correction unit 14 corrects the output prediction result in accordance with an input from the input unit 30.

[0076] Here, the input unit 30 is a device capable of receiving an input of an instruction to correct an error in the prediction result predicted by the formation region prediction unit 12c. The input unit 3 may be, for example, an input key, a keyboard, or a touch panel.

[0077] The output control unit 13b causes the output device 3 to output information on the formation area predicted by the formation area prediction unit 12c and the prediction result corrected by the prediction result correction unit 14. The output control unit may output the corrected prediction result every time the prediction result is corrected by the prediction result correction unit 14.

[0078] <Storage section 20b> The storage unit 20b will now be described. Image data 21, a prediction model 22, and prediction / correction results 23 may be stored in the storage unit 20b.

[0079] When the formation region prediction section 12c is replaced with the first formation region prediction section 12a and the second formation region prediction section 12b, the prediction model 22 may be replaced with a first prediction model 22a and a second prediction model 22b.

[0080] The prediction / correction result 23 is information about the area predicted by the formation area prediction unit 12c and information about the formation area corrected by the prediction correction result unit 14. When the prediction result is corrected, the prediction / correction result 23 may store only the prediction result after correction, or may store both the prediction results before and after correction.

[0081] (Processing performed by information processing device 1b) An overview of an information processing method according to one embodiment of the present invention will be described with reference to Fig. 9. Fig. 9 is a flowchart showing an overview of an information processing method performed by the information processing device 1b.

[0082] Steps S1 to S3 are as described above.

[0083] In the correction instruction receiving step S4, the prediction result correction unit 14 receives an input of a correction instruction for the prediction result output by the output control unit 13b. In step S4, the input unit 30 allows input of a correction of the prediction result.

[0084] In a correction step S5, the prediction result correction unit 14 corrects the prediction result in accordance with the correction instruction input in step S4. In an output control step S6, the output control unit 13b causes the output device 3 to output the corrected prediction result.

[0085] The information processing device 1b may be capable of correcting an error in the output prediction result and saving the result as a corrected prediction result. In this case, training data may be created in which an image corresponding to the prediction result before correction and input to the prediction model 22 is used as an explanatory variable and the corrected prediction result is used as a target variable. This makes it possible to improve the prediction accuracy of the prediction model 22 by performing machine learning using new training data on the prediction model 22. Furthermore, according to the above configuration, the information processing device 1b can repeatedly create training data, so that the accuracy of the prediction model 22 can be continuously improved.

[0086] [Use of information processing device] The above-mentioned information processing device may be used to evaluate the characteristics of a sample surface covered with a liquid. Therefore, the information processing device and the information processing device may be suitably used to evaluate a sample used in a state covered with a liquid. An example of such a sample may be a medical device or a medical material used in an animal or human body in a state covered with an internal liquid. However, the sample is not limited to a medical device or a medical material, and may be used for other purposes.

[0087] Examples of the surface characteristics of a sample that can be evaluated in the information processing device according to an embodiment of the present invention include surface wettability, affinity with liquid, solute adsorption behavior, slipperiness, physical resistance such as unevenness, liquid retention (hydration in the case of water), swelling, and changes therein over time. Here, the evaluation of the adsorption behavior of a solute is intended to evaluate the adsorption behavior of the solute on a sample surface covered with a liquid containing the solute. When the liquid is water, for example, the retention of the liquid is mainly due to chemical factors such as hydrogen bonds and van der Waals forces acting between the molecules on the sample surface and water, and the retention (hydration in the case of water) varies depending on the sample. The evaluation of the retention of a liquid is intended to evaluate the difference in the retention near the sample surface. For example, if the hydration force is strong, it becomes difficult for the spray gas to move the water, so the formation region becomes smaller or the formation region becomes difficult to observe. In addition, in a hydrated state, the surface of a solid usually becomes slippery when the solid is moved, for example, like a ski or skate blade. Therefore, it is possible to evaluate the slipperiness because there is a correlation between the hydration force and the slipperiness. The physical resistance is intended to mean that, for example, if the unevenness of the sample surface is large, the formation area becomes small in the sense that it inhibits the movement of the injected gas, and if the unevenness is small, the formation area becomes large. In addition, the swelling property is intended to mean that, for example, when at least a part of the sample is immersed in a liquid, if the sample swells, the formation area becomes small in the sense that the swollen sample surface inhibits the movement of the injected gas. In addition, when the entire sample is immersed in a liquid, the thickness of the liquid layer between the liquid surface and the sample surface becomes thin when it swells, so that the pressure of the injected gas is easily applied to the sample surface, and the formation area becomes large. This is intended to evaluate the swelling property of the sample. Of course, it is also possible to comprehensively evaluate the sample surface by incorporating both physical and chemical factors.

[0088] Examples of the characteristics of a sample surface that can be evaluated by an information processing device according to one embodiment of the present invention include surface wettability, affinity with liquid, solute adsorption behavior, slipperiness, physical resistance such as unevenness, liquid retention (hydration power in the case of water), swelling, and changes therein over time.

[0089] For example, when evaluating wettability, the wettability of the sample surface with respect to a specific liquid can be evaluated based on information about the formation region. The wettability can be evaluated to be lower as the length of the formation region is longer, the area of ​​the formation region is larger, and the volume of the formation region is larger.

[0090] Alternatively, the wettability may be evaluated based on the amount of change in size of the formation region at two or more points in time after the gas jet. For example, the greater the amount of change in size at two or more points in time between the start and end of the gas jet, the easier it is for the liquid to move on the sample surface, so it can be evaluated as less wettable. Also, for example, the greater the amount of change in size at the end of the gas jet and two or more points in time after the end of the gas jet, the easier it is for the sample surface to be covered with liquid again, so it can be evaluated as more wettable.

[0091] The wettability can also be evaluated based on the results of continuous or time-dependent measurement of the size, in which case continuous changes in the size can be tracked, providing insight into the behavior over time that cannot be evaluated by measurements at one or more points in time.

[0092] Furthermore, the amount of gas injected, the thickness of the liquid layer covering the surface of the sample, the type of liquid, the distance between the sample surface and the gas injection port, etc. may be changed, and the evaluation may be performed based on the amount of change in the size at that time.

[0093] In addition, for example, when a liquid containing a solute whose adsorption behavior is to be evaluated is used in the measurement, the adsorption behavior of the solute on the sample surface may be evaluated. The greater the change in length of the formation region when the liquid containing the solute is used and the length of the formation region of the blank, the higher the adsorption of the solute on the sample surface can be evaluated.

[0094] Alternatively, by varying the time of immersion in a liquid containing a solute whose adsorption behavior is to be evaluated and using the amount of change as an indicator, it is possible to evaluate the relationship between the time of immersion in the liquid containing the solute and the amount of adsorption onto the sample surface.

[0095] Conventionally, a method for evaluating a sample surface based on a formation region includes a method (rule-based method) in which rules are set in advance and the characteristics of the formation region are measured by an algorithm based on the set rules. However, as a result of examining the rule-based method, the inventors have found that it is difficult to measure objects that do not follow the set rules, and in practice, measurements of such objects are often performed manually. Therefore, there are problems such as differences due to the person performing the measurement, reduced reproducibility, reduced work efficiency, and psychological burden on the operator, and due to these limitations, in practice, only a part of the acquired image data can be used for evaluation, making it difficult to obtain useful information characteristic of the sample. In other words, a new problem has been found that there is room for improvement in terms of work efficiency. Such a problem has not been found in the past in methods for evaluating sample surfaces, and is a new problem.

[0096] According to the above configuration, by preparing the above-mentioned teacher data, it is possible to automatically measure the characteristics of the formation area without setting rules in advance. According to the above configuration, it is possible to measure the characteristics of the formation area even for images of complex formation areas that cannot be measured by rule-based methods. In addition, since a prediction model learned by machine learning is used, it is possible to predict information about the formation area faster and with higher reproducibility than manual methods.

[0097] Furthermore, in evaluating the surface of a medical device or the like, it is difficult to evaluate the surface in a liquid environment in which the medical device or the like is actually used, and it is particularly difficult to properly evaluate the surface of a medical device or the like that includes a narrow curved surface, such as a guide wire or a catheter, but according to an information processing device according to an embodiment of the present invention, it is possible to properly evaluate the surface of a medical device or the like that includes a narrow curved surface in a liquid environment in which it is actually used. Therefore, an embodiment of the present invention can be used in the development and manufacture of medical devices and medical materials.

[0098] In addition, by evaluating the characteristics of a sample surface that has been subjected to a surface treatment such as coating, the information processing device according to one embodiment of the present invention can be used in the development of coating films, etc. Therefore, one embodiment of the present invention is very useful in the field of developing coating films, etc.

[0099] [Software implementation example] The control blocks of the information processing devices 1, 1a, 1b (particularly, the units included in the control unit 10) may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or may be realized by software.

[0100] In the latter case, the information processing device 1, 1a, 1b includes a computer that executes instructions of a program, which is software that realizes each function. The computer includes, for example, one or more processors, and a computer-readable recording medium that stores the program. The object of the present invention is achieved by the processor reading the program from the recording medium and executing it in the computer. The processor may be, for example, a CPU (Central Processing Unit). The recording medium may be a "non-transient tangible medium", such as a ROM (Read Only Memory), a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like. The computer may further include a RAM (Random Access Memory) that expands the program. The program may be supplied to the computer via any transmission medium (such as a communication network or a broadcast wave) that can transmit the program. Note that one aspect of the present invention may also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

[0101] 〔summary〕 The information processing device according to a first aspect of the present invention comprises an acquisition unit that acquires an image of a specimen and a liquid when gas is injected into a liquid covering the specimen, a formation region prediction unit that inputs the image into a prediction model and predicts information about a formation region formed by the movement of at least a portion of the liquid covering the specimen by the injection of the gas, and an output control unit that outputs the prediction result by the formation region prediction unit to an output device, and the prediction model is trained by machine learning using training data including an explanatory variable including a specimen image of the specimen and the liquid when gas is injected into the liquid covering the specimen, and a target variable including information about a specimen region identified in advance by a person having a predetermined skill based on the specimen image, the specimen region being formed by the movement of at least a portion of the liquid covering the specimen.

[0102] An information processing device according to a second aspect of the present invention comprises an acquisition unit that acquires an image of the specimen and the liquid when gas is injected onto the liquid covering the specimen, a first formation region prediction unit that inputs the image into a first prediction model to predict a formation region formed by the movement of at least a portion of the liquid covering the specimen by the injection of the gas, a second formation region prediction unit that inputs region information indicating the predicted formation region into a second prediction model to predict information regarding the formation region, and an output control unit that outputs the prediction result by the formation region prediction unit to an output device, wherein the first prediction model is trained by machine learning using teacher data including an explanatory variable including a specimen image of the specimen sample and the liquid when gas is injected onto the liquid covering the specimen sample, and a target variable including region information indicating a specimen region identified in advance based on the specimen image by a person having a predetermined skill, the specimen region being formed by the movement of at least a portion of the liquid covering the specimen sample, and the second prediction model is trained by machine learning using teacher data including an explanatory variable including the region information, and a target variable including information regarding the specimen region determined in advance by a person having a predetermined skill.

[0103] The information processing device according to aspect 3 of the present invention may further include, in aspect 1 or 2, an input unit that accepts input of a correction instruction to instruct a correction to the output prediction result, and a prediction result correction unit that corrects the prediction result in response to the correction instruction.

[0104] An information processing device according to a fourth aspect of the present invention is in any one of the first to third aspects, wherein the information about the formation region and the information about the specimen region may be at least one of the dimensions, area, shape, and position of the formation region.

[0105] In the information processing device according to a fifth aspect of the present invention, in any one of the first to fourth aspects, the sample may have a columnar or cylindrical shape.

[0106] A sixth aspect of the present invention is any one of the first to fifth aspects, wherein the liquid contains a solute, and the information on the formation region may be information for evaluating an adsorption behavior of the solute on the surface of the sample.

[0107] A seventh aspect of the present invention is an information processing method including an acquisition step of acquiring an image of a specimen and a liquid when gas is injected into the liquid covering the specimen, a formation region prediction step of inputting the acquired image into a prediction model to predict information about a formation region formed by the movement of at least a portion of the liquid covering the specimen by the injection of the gas, and an output control step of causing an output device to output the prediction result by the formation region prediction unit, wherein the prediction model is trained by machine learning using training data including an explanatory variable including a specimen image of the specimen specimen and the liquid when gas is injected into the liquid covering the specimen, and a target variable including information about a specimen region identified in advance by a person having a predetermined skill based on the specimen image and formed by the movement of at least a portion of the liquid covering the specimen.

[0108] An eighth aspect of the present invention includes an acquisition step of acquiring an image of a specimen and a liquid when gas is injected onto a liquid covering the specimen, a first formation region prediction step of inputting the captured image into a first prediction model to predict a formation region formed by movement of at least a portion of the liquid covering the specimen by the injection of the gas, a second formation region prediction step of inputting region information indicating the predicted formation region into a second prediction model to predict information regarding the formation region, and an output control step of causing an output device to output a prediction result by the second formation region prediction unit, wherein the first prediction model is configured to predict the movement of at least a portion of the liquid covering the specimen. the second prediction model is trained by machine learning using teacher data including explanatory variables including a specimen image of the specimen sample and the liquid when gas is injected into the liquid, and a target variable including area information indicating a specimen area identified in advance based on the specimen image by a person having a specified skill, the specimen area being formed by the movement of at least a portion of the liquid covering the specimen sample; and the second prediction model is trained by machine learning using teacher data including explanatory variables including the area information, and a target variable including information about the specimen area determined in advance by a person having a specified skill.

[0109] A ninth aspect of the present invention is a control program for causing a computer to function as the information processing device of the first aspect, the control program causing a computer to function as the acquisition unit, the formation region prediction unit, and the output control unit.

[0110] A tenth aspect of the present invention is a control program for causing a computer to function as the information processing device of the second aspect, and is a control program for causing a computer to function as the acquisition unit, the first formation area prediction unit, the second formation area prediction unit, and the output control unit.

[0111] An eleventh aspect of the present invention is a computer-readable recording medium having the control program of the ninth or tenth aspect recorded thereon. EXAMPLES

[0112] The present invention will be described in more detail below with reference to examples and comparative examples, but the present invention is not limited to these examples.

[0113] Example 1 The error when the measurement was performed manually was compared with that when the prediction model was used. The explanatory variables were approximately 1,500 specimen images of the area formed by the movement of at least a part of the liquid present on the surface of the sample, which was generated by injecting gas onto the surface of the sample covered with liquid, and the objective variables were information on the specimen area previously identified based on these specimen images by a person with a certain level of skill. The training data was then created using the explanatory variables of approximately 1,500 specimen images of the area formed by the movement of at least a part of the liquid present on the surface of the sample, which was generated by injecting gas onto the surface of the sample covered with liquid, and the objective variables of information on the specimen area previously identified based on these specimen images by a person with a certain level of skill. The images were then shifted up, down, left, right, or zoomed, and the number of images was increased to 15,000, and each specimen image and information on the specimen area identified based on the specimen images were added as training data. These training data were input into the prediction model. Unet and Resnet were used as the prediction models.

[0114] The hydrophilicity was evaluated using guide wire A (diameter 0.038 inches) as sample A and guide wire B (diameter 0.035 inches) as sample B. For the evaluation of hydrophilicity, the sample was fixed with tape so that it was in contact with the bottom of a petri dish and immersed in 5 g of distilled water. One minute after the start of immersion, nitrogen gas was sprayed for 2 seconds at a spray volume of 5 kPa from vertically above the sample (distance between the nozzle and the distilled water surface: 10.0 mm) using a wettability evaluation device Wettio NCW-100 (manufactured by Yamato Scientific Co., Ltd.). By spraying nitrogen gas, at least a part of the liquid present on the surface of the sample was moved linearly. Ten photographs of the linear region generated on the surface of the sample were taken for each sample. The location of the linear region was inferred from the captured images using the Unet model, and then a Resnet model was generated from the results, and the coordinates of both ends of the linear region were regressively inferred to obtain a predicted value of the length of the line segment. In addition, the correct value of the length of the line segment was manually obtained using the same photograph. The results are shown in Figure 11.

[0115] 11, it can be seen that the error between the predicted value and the correct value is within about 0.16 mm. Therefore, it has been shown that by using an information processing device according to an embodiment of the present invention, it is possible to predict the area formed on the sample surface without the need for a human to set rules.

[0116] Example 2 Using the prediction model used in Example 1, the dimensions of the region were predicted from an image of the region formed by at least partial movement in the same manner as in Example 1, and the time required for prediction was compared with the time required for manual measurement. The results are shown in Table 1.

[0117] [Table 1]

[0118] As can be seen from Table 1, when the area of ​​the sample surface was predicted using the prediction model, it was possible to complete the measurement in a much shorter time than when the measurement was performed manually. Furthermore, even when the accuracy rate of the prediction model was set at 70%, the measurement was completed in a time that was sufficiently shorter than the time required to perform the measurement entirely manually. Therefore, it was demonstrated that the area of ​​the sample surface can be measured in a shorter time by using the information processing device according to one embodiment of the present invention.

[0119] Example 3 Under the same conditions as in Example 1, gas was injected onto guidewire A (diameter 0.038 inches) of sample A and guidewire B (diameter 0.035 inches) of sample B. Using the prediction model used in Example 1, the length of the line segment of the formed region was measured in images taken continuously (33 images / second) until at least a part of the liquid (distilled water) present on the surfaces of samples A and B moved and a region was formed. The results are shown in FIG. 12.

[0120] FIG. 12 shows that the length of the region on the sample surface from the start to the end of gas injection can be predicted. In addition, it was possible to measure the behavior of a wider range of regions than the parts that had been measured manually up until now. Therefore, it was shown that by using the information processing device according to one embodiment of the present invention, it is possible to measure the process of a region being formed on the sample surface, which could not be identified by conventional methods. In addition, when the results of sample A and sample B are compared, the line segment lengths of the formed region showed different values. Therefore, the information processing device according to one embodiment of the present invention can also confirm the differences in behavior between different samples. [Explanation of symbols]

[0121] R sample N Nozzle G Gas L liquid Q Liquid level P formation area

Claims

1. an acquisition unit that acquires an image of the sample and the liquid when gas is injected into the liquid covering the sample; a formation region prediction unit that inputs the captured image into a prediction model and predicts information about a formation region formed by movement of at least a portion of the liquid covering the sample due to the injection of the gas; an output control unit that outputs a prediction result by the formation region prediction unit to an output device; Equipped with The predictive model is The method is learned by machine learning using teacher data including: explanatory variables including a specimen image of the specimen and the liquid when gas is injected onto the liquid covering the specimen; and a target variable including information on a specimen region that is specified in advance based on the specimen image by a person having a predetermined skill and is formed by the movement of at least a part of the liquid covering the specimen. Information processing device.

2. an acquisition unit that acquires an image of the sample and the liquid when gas is injected into the liquid covering the sample; a first formation region prediction unit that inputs the captured image into a first prediction model and predicts a formation region that is formed by movement of at least a portion of the liquid covering the sample due to the injection of the gas; a second formation region prediction unit that inputs region information indicating the predicted formation region into a second prediction model to predict information regarding the formation region; an output control unit that outputs a prediction result by the second formation region prediction unit to an output device; Equipped with The first prediction model is The method is learned by machine learning using teacher data including: explanatory variables including a specimen image of the specimen and the liquid when gas is injected onto the liquid covering the specimen; and a target variable including area information indicating a specimen area that is specified in advance based on the specimen image by a person having a predetermined skill and is formed by the movement of at least a part of the liquid covering the specimen; The second prediction model is The method is learned by machine learning using teacher data including an explanatory variable including the region information and a target variable including information on the sample region previously determined by a person having a predetermined skill. Information processing device.

3. an input unit that receives an input of a correction instruction for correcting the output prediction result; Further comprising a prediction result correction unit that corrects the prediction result in response to the correction instruction.

3. The information processing device according to claim 1 or 2.

4. The information processing apparatus according to claim 1 , wherein the information on the formation region and the information on the specimen region are at least any one of dimensions, area, shape, and position of the formation region and the specimen region.

5. The shape of the sample is cylindrical or cylindrical.

3. The information processing device according to claim 1 or 2.

6. The liquid includes a solute; The information regarding the formation region is information for evaluating the adsorption behavior of the solute on the surface of the sample.

3. The information processing device according to claim 1 or 2.

7. acquiring an image of the sample and the liquid when gas is injected into the liquid covering the sample; a formation region prediction step of inputting the captured image into a prediction model to predict information about a formation region formed by movement of at least a portion of the liquid covering the sample by the injection of the gas; an output control step of causing an output device to output a prediction result by the formation region prediction unit; Including, The predictive model is The method is learned by machine learning using teacher data including: explanatory variables including a specimen image of the specimen and the liquid when gas is injected onto the liquid covering the specimen; and a target variable including information on a specimen region that is specified in advance based on the specimen image by a person having a predetermined skill and is formed by the movement of at least a part of the liquid covering the specimen. Information processing methods.

8. acquiring an image of the sample and the liquid when gas is injected into the liquid covering the sample; a first formation region prediction step of inputting the captured image into a first prediction model to predict a formation region formed by movement of at least a portion of the liquid covering the sample due to the injection of the gas; a second formation area prediction step of inputting area information indicating the predicted formation area into a second prediction model to predict information regarding the formation area; an output control step of causing an output device to output a prediction result by the second formation region prediction unit; Including, The first prediction model is The method is learned by machine learning using teacher data including: explanatory variables including a specimen image of the specimen and the liquid when gas is injected onto the liquid covering the specimen; and a target variable including area information indicating a specimen area that is specified in advance based on the specimen image by a person having a predetermined skill and is formed by the movement of at least a part of the liquid covering the specimen; The second prediction model is The method is learned by machine learning using teacher data including an explanatory variable including the region information and a target variable including information on the sample region previously determined by a person having a predetermined skill. Information processing methods.

9. A control program for causing a computer to function as the information processing device according to claim 1 , the control program causing a computer to function as the acquisition unit, the formation region prediction unit, and the output control unit.

10. A control program for causing a computer to function as the information processing device described in claim 2, the control program causing a computer to function as the acquisition unit, the first formation area prediction unit, the second formation area prediction unit, and the output control unit.

11. A computer-readable recording medium having the control program according to claim 9 or 10 recorded thereon.

Images