Information processing system, information processing method, and article

Abstract

An information processing system (1, 2) according to an embodiment comprises: an acquisition unit (101, 301) that acquires a captured image of the color of a coloring reagent in a coloring reaction using the feces of a subject; and a determination unit (103, 303) that, on the basis of the captured image, determines intestinal environment information pertaining to the state of the intestinal environment of the subject.

Description

Information processing system, information processing method, and article

[0001] The present invention relates to an information processing system, an information processing method, and an article.

[0002] One method for analyzing feces is to use a color reagent (dye) that undergoes a color reaction when mixed with the sample. For example, various components contained in the feces can be analyzed by visually comparing the color of the reaction solution after the color reaction with a color legend or by measuring the spectral absorbance.

[0003] International Publication No. 2020 / 097043

[0004] However, although the more color reagents used, the more versatile the analysis becomes, the more time and effort is required to visually observe and measure color changes.

[0005] An object of the present invention is to provide an information processing system, an information processing method, and an article that can easily perform fecal analysis using a color reaction.

[0006] In order to solve the above-mentioned problems and achieve the objectives, the present invention provides an information processing system comprising an acquisition unit that acquires an image of the color of a color reagent in a color reaction using the subject's stool, and a determination unit that determines intestinal environment information regarding the state of the subject's intestinal environment based on the captured image.

[0007] The present invention also provides an information processing method that includes acquiring an image of the color of a color reagent in a color reaction using the subject's stool, and determining intestinal environment information regarding the state of the subject's intestinal environment based on the image.

[0008] The present invention also provides an information processing system that includes a learning unit that performs machine learning for a plurality of subjects that provide training data using the features of the subjects and intestinal environment information regarding the state of the intestinal environment of the subjects.

[0009] According to the present invention, it is possible to provide an information processing system, an information processing method, and an article that can easily perform fecal analysis using a color reaction.

[0010] FIG. 1 is a diagram illustrating an example of the configuration of an information processing system. FIG. 2 is a diagram illustrating an example of the hardware configuration of an information processing device. FIG. 3 is a flowchart illustrating a processing procedure of the information processing system. FIG. 4 is a diagram illustrating the processing of a calculation unit. FIG. 5 is a diagram illustrating a trained model of the information processing system. FIG. 6 is a diagram illustrating an example of the configuration of an information processing system. FIG. 7 is a diagram illustrating an example of the configuration of an information processing system.

[0011] Hereinafter, an information processing system, an information processing method, and an article according to embodiments will be described with reference to the drawings. Note that the following embodiments are not limited to the following description. Furthermore, each embodiment can be combined with other embodiments or conventional techniques as long as there is no contradiction in the processing content.

[0012] In the following embodiments, a "subject" refers to a person whose feces (biological sample) is analyzed using a color reaction, and is typically a patient. However, the subject may also be a healthy individual whose health condition is being analyzed. The subject may also be an infant under one year of age, whose intestinal environment is more difficult to obtain than that of an adult, and for whom more careful health management is required.

[0013] Furthermore, the subject of analysis is not necessarily limited to humans, but may be, for example, animals such as pets, livestock, etc. When the subject of analysis includes animals other than humans, it will be referred to as a "subject" or "specimen."

[0014] Furthermore, the term "user" refers to a person who uses the information processing system according to the embodiment, and is typically a medical professional, but does not necessarily have to be a medical professional. For example, the user may be the subject himself / herself, or a close relative, friend, or acquaintance of the subject.

[0015] (Embodiment) The configuration of an information processing system 1 according to the present invention will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of the configuration of the information processing system 1. As shown in Fig. 1, the information processing system 1 according to the embodiment includes an information processing device 10 and an image capturing device 20. The information processing device 10 and the image capturing device 20 are connected to each other by any communication means. As the communication means, any network such as a LAN (Local Area Network) or a WAN (Wide Area Network) can be applied.

[0016] The information processing device 10 is a server device that provides a service for analyzing feces using a color reaction. For example, the information processing device 10 includes an acquisition unit 101, a calculation unit 102, a determination unit 103, an evaluation unit 104, and an output control unit 105. The acquisition unit 101, the calculation unit 102, the determination unit 103, the evaluation unit 104, and the output control unit 105 will be described later. The term "color reaction" used herein refers to a reaction accompanied by a color development or discoloration phenomenon. The color development or discoloration phenomenon is a phenomenon that can be detected as a difference in the wavelength of light, and refers to a phenomenon in which the values ​​of image data (e.g., at least one color score in the RGB color space, Lab color space, CMYK color space, HSV color space, and Gray color space) obtained when a sample after the color reaction is photographed using a photographing camera (e.g., Sony's Cyber-shot (DSC-WX500, WW220188)) change.

[0017] The image capturing device 20 is a device that captures an optical image (captured image) that is visible using visible light. For example, the image capturing device 20 corresponds to a smartphone equipped with a camera module capable of capturing optical images, or a camera capable of capturing optical images. For example, the image capturing device 20 captures an image of a color sensor array that has undergone a color reaction with a biological sample of the subject, and sends the captured image to the information processing device 10.

[0018] A color sensor array is a multi-well plate (e.g., a 96-well plate) in which multiple color reagents are arranged in individual wells (holes). The bottom of each well of the multi-well plate is transparent, and color changes in the color reagents in each well can be seen from the bottom.

[0019] Here, the color reagent for feces can be appropriately selected from one or more known color reagents, such as picric acid, methyl violet, malachite green, o-cresol red, thymol blue, 2,4-dinitrophenol, Congo red, methyl orange, methyl yellow, bromophenol blue, bromocresol green, bromophenol red, methyl red, litmus (blue), litmus (red), methyl purple, bromocresol purple, chlorophenol red, bromothymol blue, p-nitrophenol, neutral red, phenol red, and p-naphtholphthalein. , cresol red, phenolphthalein, thymolphthalein, alizarin yellow, 1,3,5-trinitrobenzene, nitrazine yellow, rosolic acid, bromopyrogallol red, pyrocatechol violet, 4-(4-diethylaminostyryl)-1-methylpyridinium iodide, Brooker merocyanine dye, Nile red, 2-[2-cyano-4-[(N-succinimidyloxy)carbonyl]phenyl]-1,3a,6a-triazapentalene, Reichardt's reagent, 4-dimethylamino-4-nitrostilbene , phthalocyanine, 15-crown 4 [4-(2,4-dinitrophenylazo)phenol], 18-crown 5 [4-(2,4-dinitrophenylazo)phenol], Disperse Orange 25, flavone, 3-hydroxyflavone, chalcone, cyanidin chloride, cyanine, pelargonidin chloride, peonidin, delphinidin, petunidin, malvidin, α-carotene, β-carotene, γ-carotene, δ-carotene, ε-carotene, lycopene, 9-anthryldiazomethane, DMEQ-hydrazide, rhodamine B, alcian Blue, Fluorescein, Acridine Orange Base, Pigment Blue 15, Iron(II) Phthalocyanine, Magnesium(II) Phthalocyanine, Zinc Phthalocyanine, Cobalt(II) Phthalocyanine, Dilithium Phthalocyanine, Tin(II) Phthalocyanine, Sodium Phthalocyanine, Aluminum Phthalocyanine Chloride, Lead(II) Phthalocyanine, Titanyl Phthalocyanine, Manganese(III) Tetraphenylporphyrin Chloride, 5,10,15,20-Tetraphenylporphyrinatozinc, 5,10,15,Examples of the color reagent include one or more of 20-tetrakis(2,4,6-trimethylphenyl)porphyrinatozinc, 5,10,15,20-tetrakis(pentafluorophenyl)porphyrinatozinc, and 5,10,15,20-tetrakis(4-methoxyphenyl)porphyrinatocobalt, and it is preferable to select three to five color reagents from these. Among the above color reagents, more preferable examples include one or more of thymol blue, bromocresol green, methyl red, bromocresol purple, chlorophenol red, bromothymol blue, phenol red, bromopyrogallol red, Nile red, Reichardt's reagent, rhodamine B, Alcian blue, magnesium(II) phthalocyanine, zinc phthalocyanine, sodium phthalocyanine, and phthalocyanine chloroaluminum, and it is preferable to select three to five color reagents from these. Among the above color reagents, the most preferred are one or more of bromocresol green, methyl red, Reichardt's reagent, bromocresol purple, bromopyrogallol red, bromothymol blue, phenol red, and chlorophenol red, and it is preferable to select three or more and five or less color reagents from these.

[0020] The color reagent may change color depending on the pH or polarity of the fecal-based biological sample, and more preferably changes color depending on both pH and polarity. By using such a color reagent, the concentration or abundance of intestinal metabolites, and the abundance ratio, abundance, or Shannon index of intestinal bacteria can be determined with high accuracy. Examples of color reagents that change color depending on the pH of the fecal-based biological sample include bromocresol green, methyl red, bromocresol purple, bromopyrogallol red, bromothymol blue, phenol red, and chlorophenol red. Examples of color reagents that change color depending on the polarity of the fecal-based biological sample include Reichardt's reagent.

[0021] 1 is merely an example, and the present invention is not limited thereto. For example, the image capturing device 20 does not necessarily have to be connected to the information processing device 10. Images captured by the image capturing device 20 may be stored in the information processing device 10 via any portable recording medium, such as a DVD (Digital Versatile Disc).

[0022] Furthermore, the color sensor array is not limited to a porous plate, but may also be a test strip (such as a paper device) or other article in which multiple color reagents are arranged at individual positions. The test strip can be made of any material, such as paper, plastic (resin), glass, fiber, ceramic, or metal. Examples of other articles include, but are not limited to, microfluidics, detection kits, diapers, toilet bowls, toilet paper, stool collection kits, enemas, pet litter, and pet litter sheets, with microfluidics, diapers, toilet bowls, and pet litter sheets being preferred.

[0023] Next, the hardware configuration of the information processing device 10 included in the information processing system 1 will be described with reference to Fig. 2. Fig. 2 is a diagram showing an example of the hardware configuration of the information processing device 10.

[0024] As shown in FIG. 2 , the information processing device 10 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, an auxiliary storage device 14, an input device 15, a display device 16, and an external I / F (Interface) 17.

[0025] The CPU 11 is a processor (processing circuit) that executes programs to comprehensively control the operation of the information processing device 10 and realize various functions of the information processing device 10. For example, the functions of the acquisition unit 101, calculation unit 102, determination unit 103, evaluation unit 104, and output control unit 105 of the information processing device 10 are realized by the CPU 11.

[0026] The ROM 12 is a non-volatile memory that stores various data (information written during the manufacturing stage of the information processing device 10) including a program for starting up the information processing device 10. The RAM 13 is a volatile memory that has a work area for the CPU 11. The auxiliary storage device 14 stores various data such as the program executed by the CPU 11. The auxiliary storage device 14 is configured, for example, by a hard disk drive (HDD), a solid state drive (SSD), or the like.

[0027] The input device 15 is a device that allows the operator of the information processing device 10 to perform various operations. The input device 15 is configured by, for example, a mouse, a keyboard, a touch panel, or hardware keys.

[0028] The display device 16 displays various types of information. For example, the display device 16 displays image data, model data, a GUI (Graphical User Interface) for receiving various operations from an operator, medical images, etc. The display device 16 is configured as, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, or a cathode ray tube display. Note that the input device 15 and the display device 16 may be integrated into one unit, for example, in the form of a touch panel.

[0029] The external I / F 17 is an interface for connecting (communicating) with any external device such as a server device 20 .

[0030] 2 is merely an example, and the present invention is not limited to this. For example, the hardware configuration of the information processing device 10 may be any configuration of a known computer or workstation.

[0031] The processing procedure of the information processing device 10 included in the information processing system 1 will be described with reference to Fig. 3. Fig. 3 is a flowchart showing the processing procedure of the information processing system 1. In the description of Fig. 3, Fig. 4 and Fig. 5 will be referred to as appropriate.

[0032] 3, the user performs a color reaction using the subject's stool (step S101). For example, the user brings the stool obtained from the subject into contact with a color sensor array to perform a color reaction.

[0033] Next, the user photographs the result of the color reaction of the subject (step S102). For example, the user uses the photographing device 20 to photograph the color sensor array 22 after the color reaction.

[0034] Then, the photographing device 20 transmits the photographed image to the information processing device 10 (step S103). As a result, the acquisition unit 101 of the information processing device 10 acquires the photographed image of the color sensor array 22 after the color reaction. Note that this process may be performed by any communication means or via any portable recording medium.

[0035] The calculation unit 102 of the information processing device 10 then calculates the color score of the subject based on the captured image (step S104). The color score is an example of a "feature amount" that quantifies the color of the color reagent. The color of the color reagent is represented by pixel values ​​in the RGB color space, Lab color space, CMYK color space, HSV color space, and Gray color space.

[0036] Here, the processing of the calculation unit 102 of the information processing device 10 will be described with reference to Fig. 4. Fig. 4 is a diagram for explaining the processing of the calculation unit 102.

[0037] The calculation unit 102 generates five captured images represented in the RGB color space, Lab color space, CMYK color space, HSV color space, and Gray color space from the captured image acquired by the acquisition unit 101. For example, if the captured image acquired by the acquisition unit 101 is an image represented in the RGB color space, the calculation unit 102 converts this image into each of the remaining four color spaces. For this conversion process, known image conversion techniques can be applied as appropriate.

[0038] The calculation unit 102 then identifies a region R1 in each image of each color space, in which the colors of the multiple color reagents are depicted. Specifically, the calculation unit 102 identifies the region of each well 22a of the color sensor array 22 from the captured image in which the color sensor array 22 is depicted, using pattern matching processing or the like. The calculation unit 102 then identifies a region R1 that includes a predetermined number of pixels (e.g., 50 pixels x 50 pixels) including the center of each well 22a.

[0039] The calculation unit 102 then calculates, as the color score, a statistical value of the pixel values ​​of a predetermined number of pixels included in each of the identified regions R1. For example, the calculation unit 102 calculates, as the color score, a median value of the pixel values ​​of the predetermined number of pixels included in the region R1.

[0040] Here, the color score is numerical data consisting of dimensional values ​​calculated from five images represented in five color spaces. For example, an image represented in the RGB color space has three-dimensional pixel values ​​of R, G, and B, so the median (statistical value) calculated from the image represented in the RGB color space is three-dimensional, such as (x1, x2, x3). Similarly, the median calculated from an image represented in the Lab color space is three-dimensional, such as (x4, x5, x6). The median calculated from an image represented in the CMYK color space is four-dimensional, such as (x7, x8, x9, x10). The median calculated from an image represented in the HSV color space is three-dimensional, such as (x11, x12, x13). The median calculated from an image represented in the Gray color space is one-dimensional, such as (x14). That is, the color score is 14-dimensional numerical data (x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13, x14).

[0041] In this way, the calculation unit 102 calculates a color score, which is 14-dimensional numerical data, for each well (coloring reagent).

[0042] Note that the processing content of the calculation unit 102 described above is merely an example, and the present invention is not limited thereto. For example, in the above example, the processing for calculating a color score using all five images represented in five color spaces has been described, but the present invention is not limited thereto. For example, the calculation unit 102 can calculate a color score using at least one image of the five images represented in the five color spaces, or can calculate a color score using images represented in two or more, three or more, or four or more color spaces.

[0043] Furthermore, in the above example, the process of calculating the median as the color score has been described, but the present invention is not limited to this. For example, the calculation unit 102 can calculate any statistical value, such as the mean value or the mode value, in addition to the median value, as the color score. Furthermore, the calculation unit 102 may calculate the color score by combining any statistical value, such as the median value, the mean value, or the mode value. In this case, even when the mean value is used, 14-dimensional numerical data is obtained, so when the median value and the mean value are combined, the color score becomes 28-dimensional numerical data. In other words, the calculation unit 102 may calculate at least one of the median value, the mean value, and the mode value as the color score.

[0044] Furthermore, for example, it is possible to appropriately correct the color score using data obtained by photographing the color sensor array before the color reaction is performed. Specifically, the color score can be corrected by subtracting the pixel values ​​of the image of the color sensor array 22 photographed before the color reaction from the pixel values ​​of the image of the color sensor array 22 photographed after the color reaction, and the above-mentioned color score can be calculated using the corrected pixel values. By correcting the color score in this way, it may be possible to suppress deviations in the color score that may be affected by the photographing environment.

[0045] Returning to the explanation of Figure 3, the determination unit 103 of the information processing device 10 generates a determination result by inputting the subject's color score into a trained model (step S105). This trained model is trained using the color scores and intestinal environment information of multiple providers who provide training data. In other words, the determination result is the intestinal environment information of the subject. The intestinal environment information is information regarding the state of the intestinal environment of the subject (provider).

[0046] Here, the trained model of the information processing system 1 will be described with reference to FIG. 5. FIG. 5 is a diagram for explaining the trained model of the information processing system 1. The upper part of FIG. 5 shows the processing during training of the trained model, and the lower part of FIG. 5 shows the processing during operation of the trained model. Here, the trained model is constructed in advance and stored in a predetermined storage area (e.g., ROM 12) available to the determination unit 103. The processing of the "learning unit" for constructing the trained model will be described later.

[0047] As shown in the upper part of Figure 5, during learning, machine learning is performed using, for example, color scores based on the feces of multiple donors 1 to N and intestinal environment information. Here, the donor's color score is calculated based on a color reaction using the donor's biological sample. The method for calculating the donor's color score is similar to the method for calculating the subject's color score by the calculation unit 102, and therefore will not be described here. The donor's intestinal environment information is a measurement value of the component to be analyzed in the biological sample, such as the concentration or abundance of short-chain fatty acids in the feces, the abundance ratio, abundance, or Shannon index of intestinal bacteria in the feces, etc. Among these intestinal environment information, the present invention provides useful results when the concentration or abundance of short-chain fatty acids in the feces is used. Examples of short-chain fatty acids in the feces include at least one selected from acetic acid, succinic acid, lactic acid, propionic acid, formic acid, butyric acid, isobutyric acid, valeric acid, and isovaleric acid. Examples of intestinal bacteria in the feces include the genus Bifidobacterium. Among these measurements, particularly useful results can be obtained when at least one selected from lactic acid, acetic acid, butyric acid, and propionic acid is used. Fecal short-chain fatty acids are known to be an energy source for the large intestine and an indicator of intestinal health. Measuring lactic acid and other fatty acids as management indicators is believed to contribute to anti-inflammatory effects, immune cell function promotion, and reduced risk of metabolic diseases (e.g., improved insulin sensitivity). Intestinal environment information can be measured using known measurement techniques. Furthermore, known classification models and known regression models can be appropriately applied as machine learning models.

[0048] As shown in the lower part of Figure 5, during operation, the determination unit 103 inputs the subject's color score into a trained model constructed by machine learning, causing the trained model to output a determination result indicating the subject's intestinal environment information. For example, a machine learning model trained using the concentration of acetic acid in the donor's feces outputs an estimate of the acetic acid concentration in the subject's feces as the determination result. In this way, the determination unit 103 determines the subject's intestinal environment information based on the color score.

[0049] The processing details of the determination unit 103 described above are merely examples, and the present invention is not limited thereto. For example, in the above description, the concentration of acetic acid in feces is estimated as intestinal environment information, but the present invention is not limited thereto. Any index can be applied as training data for machine learning, such as short-chain fatty acids or bacterial mass in feces that can be measured using known measurement techniques, or the Shannon index, which is a diversity index of intestinal microbiota. The Shannon index is a value that measures the diversity and balance of species within a community.

[0050] 6 can be executed at any time before the operation process is executed. The learning process can also be executed to update (additionally learn) an already generated trained model.

[0051] Returning to the description of Fig. 4, the evaluation unit 104 evaluates the health index of the subject based on the intestinal environment information (step S106). Here, the health index is an index (score) indicating the health state of the subject, and is calculated based on the intestinal environment information.

[0052] For example, it is believed that the higher the measured (estimated) abundance ratio (or abundance amount) of short-chain fatty acids or bacteria of the genus Bifidobacterium, the healthier the individual is. Therefore, the evaluation unit 104 classifies the estimated abundance ratios of short-chain fatty acids and bacteria of the genus Bifidobacterium into any number of ranks such as "A," "B," "C," etc., in descending order of estimated values, and outputs the classified ranks as health indices.

[0053] Furthermore, for example, the Shannon index of intestinal bacteria is thought to have a suitable range for the measured value (estimated value). Therefore, the evaluation unit 104 classifies the Shannon index of intestinal bacteria into an arbitrary number of ranks, such as "A" if it is within the suitable range, and "B," "C," etc., depending on the magnitude of the value outside the range, and outputs the classified rank as a health index.

[0054] In this way, the evaluation unit 104 evaluates the health index related to the subject's health based on the intestinal environment information (step S106).

[0055] Then, output control unit 105 outputs the health index (step S107). For example, output control unit 105 may display the health index on display device 16 or store it in ROM 12. Output control unit 105 may also transmit the health index to an information processing terminal that can be viewed by the user. The information processing terminal may display the health index transmitted from output control unit 105 on a predetermined display device or store it in a predetermined storage device.

[0056] The output target of the output control unit 105 is not limited to health indices, and may be, for example, the determination result of intestinal environment information. If a health index is not output, the configuration and processing of the evaluation unit 104 are not necessary.

[0057] As described above, in the information processing system 1 according to the present invention, the acquisition unit 101 of the information processing device 10 acquires a captured image of the color of a color reagent in a color reaction using the subject's feces. The determination unit 103 determines intestinal environment information related to the state of the subject's intestinal environment based on the captured image. This allows the information processing system 1 according to the present invention to easily perform fecal analysis using a color reaction. For example, the information processing system 1 can easily perform multifaceted fecal analysis using a color reagent.

[0058] Furthermore, in the information processing system 1, the calculation unit 102 calculates a feature amount that quantifies the color based on the captured image. This allows the information processing system 1 to easily quantify the color change of the color reagent.

[0059] Furthermore, the information processing system 1 can easily perform multifaceted analyses not only on human biological samples but also on biological samples obtained from animals such as pets and livestock, using color reagents appropriate for detecting components to be analyzed in the biological samples. This is thought to be particularly useful for understanding the health status of animals that cannot be interviewed.

[0060] Note that the content described in the above embodiment is merely an example, and the present invention is not limited thereto. For example, in the above embodiment, the process of calculating the color score using the captured image of the color sensor array 22 after the color reaction was described, but the captured image of the color sensor array 22 before the color reaction can also be used. For example, it is also possible to generate a difference image between the captured images before and after the color reaction, and calculate the color score from this difference image.

[0061] In the above embodiment, the functions of the information processing device 10 are provided as a server device, but the present invention is not limited to this. For example, the functions of the information processing device 10 may be provided in an information processing terminal such as a personal computer, tablet, or smartphone owned by an individual user.

[0062] (Modification) In the above embodiment, the information processing device 10 and the photographing device 20 are provided as separate devices in the information processing system 1, but this is not limiting. For example, the information processing device 10 and the photographing device 20 may be provided as an integrated device.

[0063] The configuration of an information processing system 2 according to a modified example will be described using Fig. 6. Fig. 6 is a diagram showing an example of the configuration of the information processing system 2. As shown in Fig. 6, the information processing system 2 includes an information processing device 30 that is compatible with a personal computer, tablet, smartphone, or the like that has a photographing function. The hardware configuration of the information processing device 30 is basically the same as the hardware configuration of the information processing device 10 shown in Fig. 2, and therefore description thereof will be omitted.

[0064] The information processing device 30 includes an acquisition unit 301, a calculation unit 302, a determination unit 303, an evaluation unit 304, and an output control unit 305. The configurations of the calculation unit 302, the determination unit 303, the evaluation unit 304, and the output control unit 305 shown in Fig. 6 are basically the same as the configurations of the calculation unit 102, the determination unit 103, the evaluation unit 104, and the output control unit 105 shown in Fig. 1, and therefore description thereof will be omitted.

[0065] The acquisition unit 301 has a configuration similar to that of the acquisition unit 101, and also includes an imaging unit 301A that performs the imaging function of the information processing device 30. The imaging unit 301A is a camera module (optical sensor) that can capture optical images, and has functions similar to those of the imaging device 20.

[0066] That is, the acquisition unit 301 controls the function of the imaging unit 301A to acquire a captured image of the color sensor array 22 after the color reaction. Specifically, under the control of the acquisition unit 301, the imaging unit 301A captures an image of the color sensor array on which a color reaction with the subject's biological sample has been carried out. The acquisition unit 301 generates a captured image based on the information captured by the imaging unit 301A. The generated captured image is sent to the calculation unit 302 and used for subsequent processing. Note that the function of generating a captured image may be provided by the imaging unit 301A.

[0067] (Learning Unit) The information processing system according to the present invention may also include a learning function for constructing the trained model described in the above embodiment.

[0068] The configuration of an information processing system 3 according to the present invention will be described with reference to Fig. 7. Fig. 7 is a diagram showing an example of the configuration of the information processing system 3. As shown in Fig. 7, the information processing system 3 includes an information processing device 40 and an imaging unit 50. The hardware configuration of the information processing device 40 is basically the same as the hardware configuration of the information processing device 10 shown in Fig. 2, and therefore a description thereof will be omitted. Furthermore, the configuration of the imaging unit 50 is basically the same as the configuration of the imaging device 20 shown in Fig. 2, and therefore a description thereof will be omitted.

[0069] The information processing device 40 is a server device that constructs a trained model for determining a subject's health index. For example, the information processing device 40 includes an acquisition unit 401, a calculation unit 402, a learning unit 403, and an output control unit 404. The configurations of the acquisition unit 401, the calculation unit 402, and the output control unit 404 shown in FIG. 7 are basically the same as the configurations of the acquisition unit 101, the calculation unit 102, and the output control unit 105 shown in FIG. 1, and therefore description thereof will be omitted.

[0070] Here, the learning unit 403 performs machine learning using the color scores based on the feces of multiple providers 1 to N and the intestinal environment information to construct a trained model for determining the intestinal environment information of the subject. For example, the learning unit 403 performs the machine learning described in the upper part of Figure 5 to generate a trained model that outputs a determination result indicating the intestinal environment information of the subject when the color score of the subject is input. The learning unit 403 stores the generated trained model in any storage area available to the determination unit 103.

[0071] That is, the learning method according to the present invention performs machine learning for a plurality of providers who provide training data, using the provider's feature amounts and the provider's intestinal environment information.

[0072] Although the case where the information processing device 40 includes the learning unit 403 has been described here, the present invention is not limited to this. For example, the learning unit 403 may be included in the information processing device 10.

[0073] Furthermore, when generating a trained model that outputs a judgment result indicating the health index of animals such as pets and livestock, it is preferable that the living organisms that provide the biological samples for training are animals rather than humans. When the living organisms that provide the biological samples for training include animals other than humans, they are referred to as "subjects (experimental subjects)."

[0074] According to the embodiment and modified examples described above, feces analysis using a color reaction can be easily carried out.

[0075] The programs executed by the information processing systems 1, 2, and 3 according to the above-described embodiments and modifications may be provided by being recorded in an installable or executable file format on a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, a DVD, or a USB (Universal Serial Bus), or may be provided or distributed via a network such as the Internet. Also, various programs may be provided by being pre-installed in a non-volatile storage medium such as a ROM.

[0076] The above-described embodiment can be arbitrarily combined with the above-described modified examples, and the above-described modified examples may be arbitrarily combined with each other.

[0077] [Examples] The present invention will be described in more detail below based on examples, but the present invention is not limited thereto. In these examples, a machine learning model was constructed corresponding to the reagent conditions, variable factors, and target variables corresponding to Examples 1 to 16 shown in Tables 1, 2, and 3, respectively, and the coefficient of determination was evaluated as an index of the estimation accuracy. Tables 1, 2, and 3 show the results of the reagent conditions, variable factors, color space, target variable, and coefficient of determination (R2 value) for each example.

[0078] In this example, 100 specimens collected from infants within one year of birth were used as human feces (biological samples).

[0079] The color reagents used in this example are as follows: CR: Chlorophenol Red (Tokyo Chemical Industry Co., Ltd. (TCI)) dissolved in ethanol at a concentration of 1 mM. BB: Bromothymol Blue (Fujifilm Wako) dissolved in ethanol at a concentration of 1 mM. BG: Bromocresol Green (Fujifilm Wako) dissolved in ethanol at a concentration of 1 mM. BP: Bromocresol Purple (Fujifilm Wako) dissolved in ethanol at a concentration of 1 mM. MR: Methyl Red (Fujifilm Wako) dissolved in ethanol at a concentration of 1 mM. PR: Phenol Red (Fujifilm Wako) dissolved in ethanol at a concentration of 0.8 mM. BR: Bromopyrogall Red (Tokyo Chemical Industry Co., Ltd. (TCI)) dissolved in ethanol at a concentration of 0.6 mM. RDE: Reichardt's reagent (Sigma) dissolved in ethanol at a concentration of 2 mM. RDM: Reichardt's reagent (Sigma) dissolved in methanol (Fujifilm Wako) at a concentration of 2 mM. RDI: Reichardt's reagent (Sigma) dissolved in isopropyl alcohol (Fujifilm Wako) at a concentration of 2 mM.

[0080] The procedure for the color reaction using feces is described below. Samples frozen at -80°C were thawed in a 37°C water bath. 20-25 mg of the sample was weighed into a 2 mL Eppendorf tube, and 24 μL of PBS was added per mg of sample weight. The sample was then vortexed for 30 seconds. The sample was then centrifuged at 9,000 G for 10 minutes, and the supernatant was collected. For samples other than RDE, RDM, and RDI, 30 μL of the supernatant was mixed with 30 μL of the color reagent on a 96-well plate, shaken for 1 minute, and then photographed with a camera. For RDE, RDM, and RDI, 54 μL of the supernatant was mixed with 6 μL of the color reagent, shaken for 1 minute, and then photographed with a camera.

[0081] To obtain color scores, the well portion of the 96-well plate was cropped to 50 pixels x 50 pixels from the image obtained above, and the resulting image was then processed using Pillow (ver. 9.4.0) in Python (ver. 3.9.18, python.org.) to obtain the mean, median, and mode values ​​of 2500 pixels for the RGB, HSV, Lab, CMYK, and gray color spaces.

[0082] Measurement of fecal gut bacterial metabolites (short-chain fatty acids) was performed as follows. Lyophilized fecal samples were disrupted using a bead mill (Shake Master, Biomedical Science) by vigorous shaking (1,500 g, 10 min) with 3.0 mm zirconia beads. 10 mg of the disrupted fecal sample was suspended in 1,000 μL of crotonic acid (an internal standard), followed by the addition of 500 μL of hydrochloric acid and 2,000 μL of ether. The tube was then vigorous shaken (1,500 g, 10 min) in the bead mill and subsequently centrifuged at 10,000 g for 10 min. After centrifugation, 80 μL of the upper ether layer was collected and added to an Agilent crimp-cap vial containing 16 μL of N-tert-butyldimethylsilyl N-methyltrifluoroacetamide. The vial was sealed and heated in an 80°C water bath for 20 min. After heating, the mixture was left overnight at room temperature to t-butyldimethylsilylate the hydroxyl groups. Calibration was performed using standard solutions (formic acid, acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid, valeric acid, lactic acid, and succinic acid) derivatized in the same manner as the test samples. The derivatized samples were analyzed using an Agilent Technologies 7890 Series GC-MS system equipped with a DB-5ms column (0.25 mm x 30 m x 0.25 μm; Agilent Technologies). Measurements were performed using Agilent MassHunter Workstation Data Acquisition software (version 10.0, Agilent Technologies), and the acquired data were analyzed using Agilent MassHunter Quantitative Analysis software (version 10.1, Agilent Technologies).

[0083] This section describes a method for analyzing the intestinal microbiota. First, enterobacterial DNA (deoxyribonucleic acid) was extracted from fecal samples using a 96 MagBead DNA / RNA Kit (Zymo Research). This DNA was amplified by polymerase chain reaction (PCR) using primers designed to cover the variable region, v1-v2, of the 16S rRNA gene (27Fmod: 5'-agrgtttgatymtggctcag-3', 338R: 5'-tgctgcctcccgtaggagt-3'). Then, 250 bp paired-end sequencing was performed using Illumina's Miseq®. The resulting gene sequences were merged, filtered, trimmed, and clustered using the Qiime2 (ver. 2021.11) Pipeline to obtain amplicon sequence variants (ASVs). Each ASV was taxonomically classified using the Silva-138-99 database, and bioinformatic bacterial flora analysis was performed.

[0084] The procedure for the color reaction in this example will be described below.

[0085] We will now explain the machine learning evaluation. All combinations of reagent conditions were evaluated using double cross-validation. For double cross-validation, the training data and test data of all samples were divided into 9:1 ratios. The training data was further divided into 5 parts, and training was performed using 4 parts, followed by testing using the remaining 1 part to calculate the coefficient of determination. This was done for all 5 combinations while changing the hyperparameters, and the hyperparameters were determined. Using the determined hyperparameters, training was performed using the training data and testing was performed using the test data. This was done for all 10 combinations, and the coefficient of determination R^2 was calculated and the median was evaluated.

[0086] Table 1 shows the verification results regarding the color reagents and color spaces.

[0087]

[0088] "Reagent conditions" indicates the type of color reagent used in the color reaction with feces. Note that Table 1 lists only a portion of the combinations of multiple color reagents, but in fact, a machine learning model was constructed using all combinations of color change data for all color reagents (3,628,800 combinations), and its accuracy was verified.

[0089] The "change factor" indicates the factor that causes the color reagent used to change color, and corresponds to at least one of "pH" and "polarity." Factors that contribute to pH include short-chain fatty acids, lactic acid, ammonia, hydrogen sulfide, bilirubin metabolites, bile acids, hydrogen gas, and methane gas. Factors that contribute to polarity include short-chain fatty acids, bile acids, lipid metabolites, amino acids, amino acid metabolites, polysaccharides, oligosaccharides, and polyphenols.

[0090] "Color space" is an item related to the number of types of color spaces used to calculate the color score. "Five types" indicates that the color score is calculated using five types of color spaces: RGB color space, Lab color space, CMYK color space, HSV color space, and Gray color space. "One type" indicates that the color score is calculated using only the RGB color space.

[0091] The "target variable" is an item that indicates the components in the biological sample being analyzed. Specifically, it corresponds to at least one of the following: metabolites produced by intestinal bacteria in feces, the Bifidobacterium genus, and the Shannon index. The "metabolites" correspond to at least one of acetic acid, succinic acid, lactic acid, propionic acid, formic acid, butyric acid, isobutyric acid, isovaleric acid, and valeric acid. The "Bifidobacterium genus" corresponds to the abundance ratio or abundance of bacteria of the Bifidobacterium genus. The "Shannon index" corresponds to the Shannon index of intestinal bacteria (a diversity index of intestinal microbiota).

[0092] "R value in regression" is an item that indicates the evaluation result of the regression model and corresponds to the value of the coefficient of determination R^2. An R value of 0.5 or more is represented as "A," an R value of 0.3 or more but less than 0.5 is represented as "B," and an R value less than 0.3 is represented as "C."

[0093] As shown in Table 1, in Examples 1 and 1', a machine learning model was constructed to predict the amount of acetic acid in feces using five types of color space, and the prediction results were evaluated using the coefficient of determination R2 value. Verification was performed for all combinations of color reagents, and particularly good evaluation results were obtained. In Table 1, the reagent conditions for Example 1 were a combination of BG, MR, RDE, and RDI, and the reagent conditions for Example 1' were a combination of PR and RDE. The R2 values ​​for both Examples 1 and 1' were rated "A," but Example 1 had particularly good evaluation results.

[0094] On the other hand, compared to Example 1, Example 2, which used one reagent condition (RDE), Example 3, which used 10 reagent conditions (CR, BB, BG, BP, MR, PR, BR, RDE, RDM, and RDI), Example 4, which combined reagents based on pH alone, Example 5, which combined reagents based on polarity alone, and Example 6, which used only the RGB color space, all had an R2 value of "B." The results of Table 1 indicate that the conditions of Example 1 are the most suitable combination of conditions for predicting (estimating) acetic acid in feces by machine learning.

[0095] Tables 2 and 3 verify the optimal color reagent combinations for each of the other analytes. By determining the evaluation results using machine learning, multiple reagent combinations can be selected for analysis. As a result, when measuring multiple analytes, optimizing the reagent combination allows multiple analytes to be measured with the minimum number of reagents. Furthermore, depending on the desired detection accuracy and the performance of the imaging device (for example, selecting from several reagent combinations with an "A" rating), it is possible to reduce costs by selecting the number of inexpensive reagents.

[0096]

[0097]

[0098] As shown in Tables 2 and 3, Examples 7 to 16' were obtained by aligning the change factors and color space conditions with those of Example 1, examining all combinations of color reagents for other analytes, examining combinations of reagent conditions with good R2 values, and identifying major combinations. Examples 7 to 16' listed in Tables 2 and 3 show combinations of color reagents that provided good evaluation results for each analyte (target variable).

[0099] REFERENCE SIGNS LIST 1, 2, 3 Information processing system 10, 30 Information processing device 11 CPU 12 ROM 13 RAM 14 Auxiliary storage device 15 Input device 16 Display device 17 External I / F 20, 50 Imaging device 101, 301, 401 Acquisition unit 102, 302, 402 Calculation unit 103, 303 Determination unit 104, 304 Evaluation unit 105, 305, 404 Output control unit 403 Learning unit

Claims

1. An information processing system comprising: an acquisition unit that acquires a captured image of the color of a color reagent in a color reaction using the subject's stool; and a determination unit that determines intestinal environment information regarding the state of the subject's intestinal environment based on the captured image.

2. The information processing system according to claim 1, further comprising a calculation unit that calculates a feature amount that quantifies the color based on the captured image.

3. The information processing system of claim 2, wherein the determination unit determines the intestinal environment information of a subject by inputting the features of the subject into a trained model trained using the features of the subject and intestinal environment information regarding the state of the intestinal environment of the subject for a plurality of subjects providing training data.

4. The information processing system according to claim 1, wherein the determination unit determines, as the intestinal environment information, at least one of the concentration or abundance of intestinal metabolites, and the abundance ratio or abundance of intestinal bacteria, and the Shannon index.

5. The information processing system according to claim 4, wherein the determination unit determines the concentration or abundance of at least one of acetic acid, succinic acid, lactic acid, propionic acid, formic acid, butyric acid, isobutyric acid, valeric acid, and isovaleric acid as the intestinal metabolite.

6. The information processing system according to claim 4, wherein the determination unit determines at least one of the abundance ratio or abundance of bacteria of the genus Bifidobacterium as the intestinal bacteria, and the Shannon index.

7. The information processing system according to claim 1, wherein the subject is an infant under the age of one.

8. The information processing system according to claim 1, wherein the color reagent is a reagent whose color changes depending on the pH or polarity of the feces-based biological sample.

9. The information processing system according to claim 1, wherein the colors are represented by pixel values ​​in each of an RGB color space, a Lab color space, a CMYK color space, an HSV color space, and a Gray color space.

10. The information processing system according to claim 1, wherein the color reagents are 3 to 5 types of color reagents.

11. The information processing system according to claim 1, wherein the color reagents are three to five of the following color reagents: bromocresol green, methyl red, Reichardt's reagent, bromocresol purple, bromopyrogallol red, bromothymol blue, phenol red, and chlorophenol red.

12. The information processing system according to claim 1, further comprising an evaluation unit that evaluates a health index related to the subject's health based on the intestinal environment information.

13. The information processing system according to claim 1, wherein the acquisition unit acquires the captured image in which the color of the color reagent after the color reaction is carried out is captured using a color sensor array in which each of a plurality of color reagents is arranged at an individual position.

14. An information processing method comprising: acquiring an image of the color of a color reagent in a color reaction using the subject's stool; and determining intestinal environment information regarding the state of the subject's intestinal environment based on the image.

15. The information processing method according to claim 14, further comprising evaluating a health index relating to the subject's health based on the intestinal environment information.

16. The information processing method according to claim 14, wherein the process of acquiring the captured image acquires the captured image using a color sensor array in which a plurality of color reagents are each arranged at an individual position.

17. The information processing method according to claim 14, wherein the color reagents are three to five of the following color reagents: bromocresol green, methyl red, Reichardt's reagent, bromocresol purple, bromopyrogallol red, bromothymol blue, phenol red, and chlorophenol red.

18. An article comprising a color reagent for use in a color reaction using a subject's stool, wherein the color reagent is one or more of bromocresol green, methyl red, Reichardt's reagent, bromocresol purple, bromopyrogallol red, bromothymol blue, phenol red, and chlorophenol red.

Images