JP2024018856A - Beauty level estimation system, beauty level estimation method, comprehensive health estimation system, and comprehensive health estimation method - Google Patents

Abstract

[Problem] The present invention provides a beauty level estimation system and beauty level that calculates the beauty level of a pet kept by a user based on the correlation between diversity data regarding the diversity of intestinal flora of the animal and the beauty level. The purpose is to provide an estimation method. [Solution] An acquisition unit that acquires the diversity data of the intestinal flora of a pet pet kept by a user, the diversity data of the intestinal flora, the diversity data of the intestinal flora of the pet, and beauty. A beauty level estimation system comprising: a calculation unit that calculates a beauty level of a pet kept by a user based on a correlation with the level. [Selection diagram] Figure 25

Description

The present invention relates to a beauty level estimation system, a beauty level estimation method, a comprehensive health estimation system, and a comprehensive health estimation method, and more specifically, data regarding the diversity of animal intestinal flora (hereinafter referred to as "diversity data"). A beauty level estimation system and a beauty level estimation method for calculating the beauty level of a pet kept by a breeder (hereinafter referred to as "user") based on the correlation between the beauty level and the beauty level; The present invention relates to a comprehensive health level estimation system and a comprehensive health level estimation method that calculates a comprehensive health level based on insurance risks presented based on the correlation between diversity data and insurance risks.

Pet animals such as dogs, cats, and rabbits, and livestock such as cows and pigs are irreplaceable to humans. In recent years, while the average lifespan of animals raised by humans has increased significantly, animals are more likely to suffer from some kind of injury or disease during their lifetime, and the increasing medical costs borne by their owners have become a problem.

In order to maintain the health of animals, it is important to manage their physical condition through daily diet and exercise, and to quickly respond to complaints. The reality is that breeders only notice that their animals are suffering from the disease when the disease progresses and some externally observable signs appear. Additionally, when the musculoskeletal system is weakened due to calcium or vitamin deficiencies, there are rarely any external signs, and the reality is that abnormalities are only noticed when the animal suffers an injury such as a fracture. .

Furthermore, health is closely related to beauty, and healthy animals, including humans, have glossy coats and are at an appropriate weight (neither thin nor fat), but if you raise them yourself, Because it is rare to carefully compare animals that live in captivity with those that are not kept in captivity, it is difficult to determine superiority or inferiority in terms of coat luster, etc. Particularly in the case of animals such as dogs and cats, whose coats and physiques vary greatly depending on their breed, it is extremely difficult for owners to determine the superiority or inferiority of their coats. If it is possible to easily determine the beauty level of a pet animal, it is expected that it will provide an opportunity for users to reconsider aspects of their pet's lifestyle, such as their pet's eating habits.

Therefore, there is a need for a means of presenting the beauty level of an animal, such as the luster of its coat, in a simple manner.

Patent Document 1 describes the effect of effectively adjusting or improving the intestinal flora by multiplying bacteria of the phylum Bacteroidetes and decreasing bacteria of the phylum Firmicutes in the intestinal flora. Although a composition for adjusting or improving intestinal flora having the following is disclosed, it does not disclose a method for presenting a beauty level from data regarding an animal's intestinal flora.

International publication 2017/094892 pamphlet

Therefore, an object of the present invention is to provide a system and method that can present the beauty level of a pet animal in a simple manner.

The present inventors analyzed and considered the diversity data of animals enrolled in pet insurance and the beauty level of the animals, such as the quality of their coat luster. We discovered that it is possible to predict the beauty level of each animal.

Furthermore, the present inventors believe that beauty is suitable for determining whether one is not sick but not healthy, that is, whether or not one is suffering from a disease, so by combining the beauty level with insurance risk, the overall health We found that it is possible to predict

That is, the present invention includes the following [1] to [12].
[1] An acquisition unit that acquires the diversity data of the intestinal flora of the pet pet kept by the user, the diversity data of the intestinal flora, the diversity data of the intestinal flora of the pet, and the beauty level. A beauty level estimation system comprising: a calculation unit that calculates the beauty level of a pet kept by a user based on the correlation between the two.
[2] The beauty level estimation system according to [1], wherein the acquisition unit further acquires basic information of a pet kept by the user.
[3] The beauty level estimation system according to [1] or [2], wherein the beauty level is a level related to hair luster or a level related to body shape.
[4] The beauty level estimation system of [1], wherein the pet animal's intestinal microflora diversity data is Shannon index.
[5] The calculation unit corrects the beauty level calculated based on the pet animal's basic information in accordance with the beauty variation factor derived using the pet animal's intestinal flora diversity data. , the beauty level estimation system of [2], which calculates a beauty level.
[6] The beauty level estimation system of [2], wherein the basic information of the pet is at least the breed and age.
[7] An acquisition unit that acquires the diversity data of the intestinal flora of the pet animal kept by the user, and a trained model to calculate the diversity data of the pet animal from the diversity data of the intestinal bacteria acquired by the acquisition unit. A beauty level estimation system comprising: a calculation unit that calculates a beauty level, wherein the trained model has learned the relationship between the diversity data of the intestinal flora of a pet animal and the beauty level of the pet animal. A beauty level estimation system characterized by being a trained model.
[8] An acquisition unit that acquires diversity data of intestinal flora of pet animals kept by the user;
a calculation unit that calculates the beauty level of a pet pet kept by the user based on the intestinal flora diversity data and the correlation between the intestinal flora diversity data of the pet animal and the beauty level; a risk calculation unit that calculates the insurance risk of the pet pet kept by the user based on the intestinal flora diversity data and the correlation between the intestinal flora diversity data of the pet animal and the insurance risk; , a health level calculation unit that calculates a health level based on beauty level and insurance risk, and a comprehensive health level estimation system.
[9] A step of obtaining diversity data of the pet animal's intestinal flora, and a step in which the computer acquires the diversity data of the pet animal's intestinal flora based on the diversity data and the correlation between the diversity data of the pet animal's intestinal flora and the beauty level. A beauty level estimation method comprising, in order, the steps of: calculating the beauty level of a pet pet kept by a user;
[10] The beauty level estimation method according to [9], further comprising a step of acquiring basic information of a pet kept by the user.
[11] The beauty level estimation method according to [10], wherein the basic information of the pet is at least breed and age.
[12] Calculating the beauty level of the pet kept by the user based on the diversity data of the pet kept by the user and the correlation between the diversity data of intestinal bacteria of the pet and the beauty level The beauty level estimation method according to [10] or [11], further comprising the step of correcting the beauty level derived in [10] or [11] based on information on food eaten by a pet kept by the user.

According to the present invention, it is possible to provide a beauty level estimating system, a beauty level estimating method, a comprehensive health level estimating system, and a comprehensive health level estimating method that can present the beauty level of a pet in a simple manner.

1 is a schematic diagram of an insurance premium calculation system. FIG. 2 is a diagram illustrating an insurance premium calculation method. It is a model diagram showing the correlation between diversity data (diversity index) and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. It is a figure showing the result of correlation analysis. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. It is a figure showing the result of correlation analysis. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. FIG. 3 is a diagram showing the correlation between diversity data and insurance risk. It is a figure showing the relationship between the number of foods and diversity data. It is a figure showing the relationship between the number of foods and diversity data. FIG. 1 is a schematic diagram of a beauty level estimation system. It is a figure explaining the beauty level estimation method. It is a figure showing the correlation between diversity data and beauty level (hair luster). It is a figure showing the correlation between diversity data and beauty level (hair luster). It is a figure showing the correlation between diversity data and beauty level (standard weight).

First, an insurance premium calculation system that is similar to the beauty level estimation system will be described. The insurance premium calculation system, together with the beauty level estimation system described below, is a necessary component for constructing a comprehensive health level estimation system.

[Insurance premium calculation system]
The insurance premium calculation system of the present invention includes an acquisition unit that acquires diversity data of intestinal bacterial flora, and a data collection unit that acquires diversity data of pets kept by the user based on the diversity data of the pets kept by the user and the correlation. A calculation section for calculating insurance risk. In the insurance premium calculation system of the present invention, it is preferable that the acquisition unit further acquires basic information about a pet kept by the user. Moreover, it is preferable that the insurance premium calculation system of the present invention further includes a setting unit that sets a correlation between the intestinal microbiota diversity data of the pet and the insurance risk.

[System overview]
FIG. 1 is a diagram illustrating an overview of an insurance premium calculation system according to an embodiment of the present disclosure. As shown in FIG. 1, the insurance premium calculation system according to this embodiment includes a server 1 and a user terminal 2. Server 1 and terminal 2 are connected via a network. The server 1 includes a processing unit (CPU) 10, a storage unit 20, and an interface unit 30. Note that users in the present invention include not only owners of pet animals, but also agents, breeders, and the like. Pet animals include dogs, cats, rabbits, birds, reptiles, and amphibians, with dogs, cats, and rabbits being preferred.

Further, the processing calculation unit (CPU) 10 includes a calculation unit 11 and a prediction calculation unit 12, the storage unit 20 includes at least a setting unit 21, and the interface unit 30 includes an acquisition unit 31 and an output unit 32. be done. The setting unit 21 may be configured to store formulas, functions, tables, or software related to the correlation between intestinal flora diversity data and insurance risk. In the case of such a configuration, the calculation unit 11 calls the formula, function, table, or software stored in the setting unit 21 based on the basic information of the pet animals kept by the user, and calculates the variety of pets kept by the user. Insurance risk can be calculated based on gender data. Note that insurance risk is an index indicating the high probability of occurrence of an event for which an insurance claim will be paid. Events for which insurance claims are payable are injuries and diseases under animal insurance. Insurance risk is reflected in insurance premiums; the higher the risk, the higher the premium, and the lower the risk, the lower the premium. Note that the insurance premium is money that the user pays to the insurance company, and the insurance money is the money that the insurance company pays to the user when an event occurs.

Here, the calculation unit 11 calculates the insurance risk of the pet kept by the user based on the diversity data of the pet kept by the user and the correlation. The embodiment of FIG. 1 further includes a prediction calculation section 12. The prediction calculation section 12 predicts the future insurance risk of the pet pet kept by the user based on the information on the food given to the pet. Calculate insurance risk by making informed adjustments.
Further, the setting unit 21 sets the correlation between the diversity data of the intestinal flora of the pet and the insurance risk.
Further, the acquisition unit 31 acquires the diversity data of intestinal flora, preferably basic information of the pet kept by the user, and the output unit 32 outputs the insurance risk, insurance premium, and information calculated by the calculation unit 11. Send the burden amount etc. to the user.

[Diversity data]
Diversity data is data related to the diversity of bacteria in the intestinal flora of animals. The high diversity of the intestinal flora means that the intestinal flora contains a wide range of different types of bacteria evenly. There are several types of diversity data indicators, so-called diversity indexes, and in the present invention, any known one may be used. Diversity indexes include the Shannon-Wiener diversity index (hereinafter sometimes abbreviated as "Shannon index"), the Simpson index, the number of unique sequences detected by a sequencer (amplicon sequence variant: ASV), and the OTU (operational taxonomic unit) number, Faith's PD, Pielou's evenness, etc.

[Measurement of diversity data]
To measure the diversity data of the intestinal flora, known metagenomic analysis methods such as amplicon sequencing and shotgun sequencing using a sequencer such as NGS or bacterial flora analysis methods can be used. For example, a method in which a sample such as feces is collected from an animal, and the DNA and RNA base sequence information of all the organisms contained in the sample is analyzed using a next-generation sequencer, thereby identifying the organisms contained in the sample. can be mentioned. Preferably, all or part of the 16S rRNA gene contained in the sample is amplified as necessary, sequenced, and the obtained sequence is analyzed using software to obtain composition data of bacteria in the sample. There are several methods.

An example of 16S rRNA gene amplicon analysis (meta-16S analysis) using NGS (next generation sequencer) will be specifically explained. First, DNA is extracted from a sample using a DNA extraction reagent, and the 16S rRNA gene is amplified from the extracted DNA by PCR. Thereafter, the base sequences of the amplified DNA fragments are comprehensively determined using NGS, low-quality reads and chimeric sequences are removed, and then the sequences are clustered to perform OTU (Operational Taxonomic Unit) analysis. An OTU is an operational classification unit for treating sequences that have a certain degree of similarity (for example, 96 to 97% or more homology) as one bacterial species. Therefore, the number of OTUs represents the number of bacterial species constituting the bacterial flora, and the number of reads belonging to the same OTU is considered to represent the relative abundance of that species. Furthermore, by selecting a representative sequence from among the number of reads belonging to each OTU, it is possible to identify the family name and genus/species name by searching the database. In this way, the number of bacterial species belonging to a particular family or phylum can be determined. Furthermore, analysis using ASV (Amplicon Sequence Variant) is also possible. Because ASVs are created after removing erroneous sequences generated during PCR and sequencing, single-base sequence variations can be distinguished, allowing for more detailed identification. The insurance premium calculation system of the present invention may use diversity data measured in advance, or may receive a fecal sample from a user, measure diversity data, and use the diversity data. When using pre-measured diversity data, the user, for example, sends a fecal sample to an intestinal bacteria analysis company, requests measurement of the intestinal flora, and receives the diversity data from the requested analysis company. receive. The user can transmit the diversity data thus received to the insurance premium calculation system of the present invention through the user terminal. Further, the configuration may be such that the requested analysis company sends the diversity data to the insurance premium calculation system of the present invention on behalf of the user.

Next, technical features of the present invention will be explained in detail.
[Calculation of insurance risk based on diversity data]
An example of a method for calculating insurance risk based on diversity data will be described below with reference to FIG. 3 (model diagram). FIG. 3 is a graph for explaining an example of an insurance risk calculation method based on diversity data according to the present embodiment. The vertical axis of this graph is insurance risk, and the horizontal axis is diversity data. Specifically, when the diversity index (Shannon index) is 3.5 (D1), the insurance risk (R1) is 1. This means that the probability that a pet animal with a diversity index of 3.5 will suffer from an injury or disease is the same as the average of all pet animals. In other words, if the average treatment cost for all pets is 5,000 yen/month, the treatment cost (paid by the owner) for a pet with a diversity index of 3.5 will also be 5,000 yen/month. means. On the other hand, when the diversity index is 3 (D2), the insurance risk (R2) is 1.2. This means that the treatment cost for a pet with a diversity index of 3.5 would be 6,000 yen.

To explain more specifically, FIG. 8 shows the insurance risk of toy poodles aged 0 to 3 years, with the vertical axis representing the insurance risk and the horizontal axis representing the diversity index (Shannon index). In addition, in order to perform statistical analysis, the diversity index was divided into four categories: (1) 2.0 or more, less than 3.0, (2) 3.0 or more, less than 4.0, (3) 4 (4) 5.0 or more, less than 6.0). As shown in Figure 8, it can be seen that there is a correlation between diversity data and insurance risk for toy poodles aged 0 to 3 years.

[Settings section]
The server or the storage unit may include a setting unit that sets a correlation between the pet animal's intestinal flora diversity data and insurance risk. The correlation here is information indicating a correspondence between the degree of diversity data and the degree of insurance risk. Correlation may be viewed as a model (function) with diversity data as input and insurance risk as output. For example, a correlation is established by statistically processing the diversity data of a plurality of pet animals and the cost of injuries and illnesses and treatment of the pet animals. Note that the pet animal used for setting the correlation and the pet animal bred by the user who is the subject of insurance risk calculation are, in principle, different individuals.

Further, the correlation may be set based on basic information of the pet pet kept by the user. Note that basic information includes age and breed. For example, the server statistically processes the diversity data of pets having basic information similar to the basic information of the pets kept by the user, and the costs of injuries and illnesses of the pets and treatment. Set up correlation. Correlations for each category of basic information (for example, age, breed, etc.) may be constructed in advance in the server, and correlations for categories corresponding to the basic information of pet animals kept by the user may be set. The setting section does not need to set the correlation every time the system is run, and the calculation section may continue to use the correlation once set by the setting section. Alternatively, the correlation may be set each time data related to insurance risk and diversity data is sequentially updated.

By setting the correlation between diversity data and insurance risk, it becomes possible to calculate the insurance risk and insurance premium based on the diversity data of pets kept by the user.

[Insurance premium calculation method]
FIG. 2 is a diagram illustrating an overview of an insurance premium calculation method according to an embodiment of the present disclosure.
As shown in FIG. 2, the insurance premium calculation method according to the present embodiment includes a step S1 of acquiring basic information about the pet pet kept by the user and diversity data of the intestinal flora, and step S2 of setting a correlation between the diversity data of the pet animals and the insurance risk; and calculating the insurance risk of the pet animals kept by the user based on the diversity data of the pets kept by the user and the correlation. Step S3 is provided in this order.

Here, it is preferable that the method further includes a step of acquiring basic information about the pets kept by the user, and it is preferable to set the correlation based on the basic information about the pets kept by the user. By setting the correlation according to the basic information of the pet pet kept by the user, it is possible to calculate an insurance premium that is more suitable for the pet animal. It is easier to first calculate a provisional insurance premium based on basic information, then modify the provisional insurance premium by taking insurance risks based on diversity data into account and calculate the final insurance premium. preferable. In this way, when modifying the provisional insurance premium based on basic information by taking into account the insurance risk based on diversity data, it is preferable to modify the provisional insurance premium by a factor of 0.5 to 2.0. . The inventors have found that, depending on diversity data, the insurance risk of individual pet animals varies between 0.5 and 2.0 times the average. In addition, we calculate insurance premiums based on the medical costs obtained by first calculating the expected medical costs based on basic information, and then modifying the medical costs according to the insurance risk derived based on the diversity data. You can.

Furthermore, the method may further include a step of predicting future insurance risks of the pet pet kept by the user based on information on food eaten by the pet pet kept by the user. Since the food that your pet eats increases diversity, future insurance premiums can be calculated by taking food information into account.

Note that the insurance premium is generally the sum of the animal's medical expenses multiplied by the insurance company's burden rate, plus the insurance company's fees, operating costs, etc. Therefore, it is possible to predict the financial burden (medical expenses) that the user will bear in the same way that the insurance premium calculation system calculates insurance premiums.

A reference example of the insurance premium calculation system is shown below.
[Reference example 1]
(Selection of dog)
About 110,000 individuals whose intestinal microbiota diversity index was measured from fecal samples during the insurance contract period (one year) were investigated for insurance risk based on insurance claims made during the same period. Since the presence or absence of injury or disease before the collection of the fecal sample is not taken into account, individuals suffering from injury or disease before the collection of the sample are also included.

(DNA extraction from fecal sample)
Fecal samples were collected from each dog and DNA was extracted as follows.
Dog fecal samples were collected by dog owners using a fecal collection kit. The fecal samples were received and suspended in fixative (10% EtOH, 1.07% NH4Cl, 5mM EDTA, 0.09% NaN3).
Next, 200 uL of the fecal suspension and 810 uL of Lysis buffer (containing 224 ug/mL Protenase K) were added to the bead tube, and the beads were crushed using a bead homogenizer (6,000 rpm, crushing 20 seconds, interval 30 seconds, crushing for 20 seconds). Thereafter, the specimen was left to stand on a heat block at 70°C for 10 minutes to perform treatment with Proteinase K, and subsequently, the sample was left to stand on a heat block at 95°C for 5 minutes to inactivate Proteinase K. DNA was automatically extracted from the lysed specimen using Chemagic 360 (PerkinElmer) according to the Chemagic kit stool protocol to obtain 100 uL of DNA extract.

(Meta-16S RNA gene sequence analysis)
Meta-16S sequence analysis was performed using a modified illumina 16S Metagenomic Sequencing Library Preparation (version 15044223 B). First, a 460 bp region containing variable regions V3-V4 of the 16S rRNA gene was amplified by PCR using universal primers (Illumina_16S_341F and Illumina_16S_805RPCR). The PCR reaction mixture consisted of 10 uL of DNA extract, 0.05 uL of each primer (100 uM), 12.5 uL of 2x KAPA HiFi Hot-Start ReadyMix (F. Hoffmann-La Roche, Switzerland), and 2.4 uL of PCR grade water. It was prepared by For PCR, after heat denaturation at 95°C for 3 minutes, a cycle of 95°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds was repeated 30 times, and finally an extension reaction was performed at 72°C for 5 minutes. The amplified products were purified using magnetic beads and eluted with 50 uL of Buffer EB (QIAGEN, Germany). The purified amplification product was subjected to PCR using Nextera XT Index Kit v2 (illumina, CA, US) and an index was added. A PCR reaction solution was prepared by mixing 2.5 uL of amplification product, 2.5 uL of each primer, 12.5 uL of 2x KAPA HiFi Hot-Start ReadyMix, and 5 uL of PCR grade water. For PCR, after heat denaturation at 95°C for 3 minutes, a cycle of 95°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds was repeated 12 times, and finally an extension reaction was performed at 72°C for 5 minutes. The indexed amplification products were purified using magnetic beads and eluted with 80-105 uL of Buffer EB. The concentration of each amplification product was measured using a NanoPhotometer (Implen, CA, US), adjusted to 1.4 nM, and then mixed in equal amounts to form a library for sequencing. The DNA concentration of the sequencing library and the size of the amplified product were confirmed by electrophoresis, and this was analyzed by MiSeq. For analysis, 2×300 bp paired-end sequencing was performed using MiSeq Reagent Kit V3. The obtained sequence was analyzed using analysis software called QIIME2 to obtain bacterial composition data.
The sequences of the universal primers used above are as follows. This universal primer can be purchased commercially.
Illumina_16S_341F
5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG- 3'

llumina_16S_805R
5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC- 3'

According to the above method, composition data of intestinal flora was obtained for all dog breeds from 0 years old to 3 years old, and diversity index (Shannon index, ASV index) was measured using QIIME2. Since outliers were excluded, the effective number of individuals was 105,335 individuals for the Shannon index and 105,677 individuals for the ASV index.

(Confirmation of injury/illness/insurance claim)
We investigated whether there were any insurance claims made during the insurance period for all of the dog breeds listed above. As mentioned above, since the presence or absence of injury or disease before the collection of the fecal sample is not taken into account, individuals suffering from injury or disease before the collection of the sample are also included.

(Correlation diagram)
The correlation between the diversity data obtained in the above experiment and insurance risk is shown in Figures 4 and 6 to 13. The vertical axis is insurance risk, and the horizontal axis is diversity index.
Figure 5 shows the results of the correlation analysis of Figure 4 (dog breeds not limited, 0-3 years old), and it can be seen that there is a correlation between diversity data and insurance risk.
Correlation coefficient: -0.97 (negative correlation)
Approximate formula: y = -0.2078x + 1.8795,
Coefficient of determination: ^R2 = 0.9438

In order to explain that a correlation can be shown even in dogs other than those over 0 years old and under 3 years old, we investigated correlations under other conditions (Figures 14 to 16).
FIG. 14 is a correlation diagram between diversity data and insurance risk for all dog breeds (136,033 individuals) aged 0 to 7 years. These results show that there is a correlation between diversity data and insurance risk not only for children aged 0 and over and under 3, but also for children aged 4 and over.
Moreover, FIG. 15 is a diagram in which the diversity index is changed to the ASV index instead of the Shannon index (105,677 individuals). It can be seen that a correlation exists between diversity data and insurance risk even if the diversity index is changed.
Furthermore, FIG. 16 is a correlation diagram between diversity data and insurance risk for all cat species (36,312 individuals) aged 0 to 3 years old. It can be seen that there is a correlation between diversity data and insurance risk not only for dogs but also for other pet animals. Note that the diversity of cats will be explained in detail in Examples 3 and 4.

(simulation)
First of all, the average medical expenses for a dog aged 0-3, regardless of breed, is 100,000 yen per year, and an animal insurance product with 50% coverage of 70,000 yen per year (including 20,000 yen in commissions, operating costs, etc.) Assume that there exists. Here, we simulate presenting an insurance premium to Shiba Inu A, who is 2 years old and has diversity data of 4.3.
First, a correlation between "Shiba, 0-3 years old" is established based on age and breed (Figure 10). Here, since Shiba Inu A has diversity data of 4.3, the insurance risk is calculated to be approximately 0.9. Therefore, Shiba Inu A's medical expenses are predicted to be 90,000 yen. Since this is an insurance product with 50% coverage, a total of 65,000 yen is offered, which is 45,000 yen plus 20,000 yen for commissions, operating costs, etc.

[Reference example 2]
As a result of our research, when comparing Pomeranians who ate only one type of commercially available dry food with Pomeranians who ate a combination of two types of dry food, the latter had a diversity index (Shannon index) of approximately 0. 2 (4.2 → 4.4). In other words, by adding the information that a Pomeranian who had been eating only one type of dry food has started eating two types of dry food together, the diversity index increases by about 0.2, and the future insurance risk becomes 0. A decrease of about .1 can be expected (Figure 13). When applied to the above simulation, it is possible to present to the user that medical expenses will be reduced by about 10,000 yen and insurance premiums will be reduced by about 5,000 yen.

[Reference example 3]
(Cat selection)
Insurance risks were investigated based on insurance claims made during the same period for approximately 110,000 individuals whose intestinal microbiota diversity index was measured from fecal samples during the insurance contract period (one year). Since the presence or absence of injury or disease before the collection of the fecal sample is not taken into account, individuals suffering from injury or disease before the collection of the sample are also included.

(DNA extraction from fecal samples and meta-16S RNA gene sequence analysis)
Using the same method as for dogs, we obtained composition data on intestinal microbiota for all cat species aged 0 to 3 years old, and measured the diversity index (Shannon index) using QIIME2. In addition, since outliers were excluded, the effective number of individuals for the Shannon index was 36,312 individuals.

(Confirmation of injury/illness/insurance claims)
We investigated whether or not there were any insurance claims made during the insurance period for all of the above cat breeds. As mentioned above, since the presence or absence of injury or disease before the collection of the fecal sample is not taken into account, individuals suffering from injury or disease before the collection of the sample are also included.

(Correlation diagram)
FIG. 16 shows the correlation between diversity data and insurance risk obtained in the above experiment. The vertical axis is insurance risk, and the horizontal axis is diversity index.
Figure 17 shows the results of the correlation analysis of Figure 16 (cat breeds not limited, 0-3 years old), and it can be seen that a correlation exists between diversity data and insurance risk.
Correlation coefficient: -0.93 (negative correlation)
Approximate formula: y = -0.1862x + 2.0147,
Coefficient of determination: ^R2 = 0.8712

In addition, Figure 18 is a diagram showing the correlation between diversity data and insurance risk for (no restrictions on cat breeds, age 0), and Figure 19 shows the correlation for (no restrictions on cat breeds, age 1). FIG. It can be seen that the correlation holds not only for the total of 0-3 years old, but also for each age group.

Furthermore, Figure 20 shows the correlation between diversity data and insurance risk for (Scottish Fold, 0-3 years old), and Figure 21 shows the correlation for (Mixed breed, 0-3 years old). FIG. It can be seen that a correlation holds even when analyzed by variety.

Furthermore, FIG. 22 is a correlation diagram between diversity data and insurance risk for all cat species (45,400 individuals) aged 0 to 7 years old. These results show that there is a correlation between diversity data and insurance risk not only for children aged 0 and over and under 3, but also for children aged 4 and over. Furthermore, it can be seen that there is a correlation between diversity data and insurance risk not only for dogs but also for cats.

(simulation)
First of all, the average medical expenses for cats aged 0-3 years, regardless of breed, is 100,000 yen per year, and an animal insurance product with 50% coverage of 70,000 yen per year (of which 20,000 yen includes commissions, operating costs, etc.) Assume that there exists. Here, a simulation is performed in which an insurance premium is offered to Scottish Fold A, who is 2 years old and has diversity data of 5.8.
First, a correlation between "Scottish Fold, 0-3 years old" is established based on age and breed (Figure 20). Here, since Scottish Fold A has diversity data of 5.8, the insurance risk is calculated to be approximately 0.8. Therefore, Scottish Fold A's medical expenses are predicted to be 80,000 yen. Since this is an insurance product with 50% coverage, a total of 60,000 yen is offered, which is 40,000 yen plus 20,000 yen for fees, operating costs, etc.

[Reference example 4]
As a result of research by the present inventors, when comparing dogs that ate only one type of food with dogs that ate a combination of two or more types of food, the latter showed an increase in diversity index (Shannon index). Got it (Figure 23). Similar results were obtained for cats (Figure 24). In other words, by adding information that a pet that had been eating one type of food has started eating two or more types of food at the same time, the diversity index will increase according to the number of types of food that it has started eating, and it will It can be expected that the insurance risk will decrease. In other words, it is possible to present to the user how much medical costs will be reduced by feeding the user multiple foods. Note that, in the present invention, information on the brand of the food and the processing method (whether it is dry food or fresh food) is not considered. Most commercially available pet foods are comprehensive nutritional foods, and the impact of food brands and processing methods on the diversity index is compared to the impact of the number of food types on the diversity index. This is because it is small.

[Reference example 4-1]
When comparing a dog that ate only one type of food with a dog that ate a combination of three types of food, the diversity index (Shannon index) of the latter increased by about 0.025 (Figure 23). By adding the information that Shiba Inu A in Example 1, who had been eating one type of food, started eating three types of food at the same time, it is thought that the diversity of Shiba Inu A would increase from 4.3 to 4.325. By referring to the "Shiba 0-3 years old" correlation (Figure 10), it can be predicted that the insurance risk will decrease from about 0.9 to about 0.89. Therefore, by applying the above simulation, it is possible to present to the user a prediction that medical expenses will decrease by about 1000 yen.

[Reference example 4-2]
As a result of our research, when we compare cats who eat only one type of commercially available dry food with cats who eat three types of dry food, the diversity index (Shannon index) of the latter increases by approximately 0.03. (Figure 24). In other words, by adding the information that Scottish Fold A, who had been eating only one type of dry food, has started eating three types of dry food, the diversity of Scottish Fold A increases from 5.8 to 5.83. Therefore, by referring to the correlation of "Scottish Fold 0-3 years old" (Figure 20), it can be predicted that the insurance risk will decrease from 0.8 to 0.79. Therefore, by applying the above simulation, it is possible to present to the user a prediction that medical expenses will decrease by about 1000 yen.

[Beauty level estimation system]
Next, the beauty level estimation system of the present invention will be explained. The beauty level estimation system can be a component for constructing a comprehensive health level estimation system together with the insurance premium calculation system described above. When constructing a comprehensive health level estimation system, the beauty level estimation system and the insurance premium calculation system may be separate systems, or the two functions of beauty level estimation and insurance premium calculation may be provided on the same system. You may also have

The beauty level estimation system of the present invention includes an acquisition unit that acquires diversity data of intestinal microbiota, and a beauty level estimation system for a pet pet kept by the user based on the correlation between the beauty level and diversity data of the pet pet kept by the user. A calculation unit that calculates a beauty level. In the beauty level estimation system of the present invention, it is preferable that the acquisition unit further acquires basic information about a pet kept by the user. Furthermore, it is preferable that the beauty level estimation system of the present invention further includes a setting unit that sets a correlation between the pet animal's intestinal microbiota diversity data and the beauty level. Further, the beauty level estimating system of the present invention preferably includes a prediction calculation unit that predicts the future beauty level of the pet pet kept by the user based on information on food given to the pet animal.

[Acquisition part]
The acquisition unit is the same as that used in the insurance premium calculation system described above.

[Calculation section]
The calculation unit calculates the beauty level of the pet pet kept by the user based on the pet animal's intestinal flora diversity data and the correlation between the pet animal's intestinal flora diversity data and the beauty level. do. The calculation unit is composed of, for example, a calculation device such as a CPU, and performs processing, calculation, and calculation using formulas, functions, tables, or software related to the correlation between the diversity data of intestinal flora and the beauty level.

The calculation unit calculates the beauty level by correcting the beauty level calculated based on the pet animal's basic information according to the beauty variation factor derived using the pet animal's intestinal microflora diversity data. It may be calculated. For example, a temporary beauty level is calculated based on basic information such as age and breed, and then the temporary beauty level is corrected according to beauty variation factors. The beauty variable factor is a variable that is set based on the correlation between the diversity data of the pet's intestinal flora and the beauty level, and is, for example, 0.01 to 3.0, preferably 0.01 to 3.0. This is a numerical value set within the range of 1 to 2.0. As a specific example, if the beauty variation factor is 0.5, the final beauty level is calculated by multiplying the temporary beauty level (specific value in this case) by 0.5 based on the basic data. Beauty level is calculated.

[Diversity data]
Diversity data is data related to the diversity of bacteria in the intestinal flora of animals, and is the same as that used in the insurance premium calculation system described above, including the measurement method.

[Settings section]
The server or the storage unit may include a setting unit that sets a correlation between the pet animal's intestinal flora diversity data and insurance risk. The correlation here is information indicating a correspondence between the degree of diversity data and the degree of beauty level. The correlation may be understood as a model (function) that uses diversity data as input and beauty level as output. For example, a correlation is established by statistically processing diversity data of a plurality of pets and information regarding the beauty of the pets (for example, coat luster). Note that the pet animal used for setting the correlation and the pet animal bred by the user whose beauty level is to be calculated are, in principle, different individuals.

Further, the correlation may be set based on basic information of the pet pet kept by the user. Note that basic information includes age and breed. For example, the server establishes a correlation by statistically processing diversity data of pets that have basic information similar to the basic information of the pets kept by the user and information regarding the beauty of the pets. do. Correlations for each category of basic information (for example, age, breed, etc.) may be constructed in advance in the server, and correlations for categories corresponding to the basic information of pet animals kept by the user may be set. The setting section does not need to set the correlation every time the system is run, and the calculation section may continue to use the correlation once set by the setting section. Alternatively, the correlation may be set each time the data regarding the beauty level and the diversity data is sequentially updated.

[Overview of beauty level estimation system]
FIG. 25 is a diagram illustrating an overview of a beauty level estimation system according to an embodiment of the present disclosure. As shown in FIG. 25, the beauty level estimation system according to this embodiment includes a server 4 and a user terminal 5. The server 4 and the terminal 5 are connected via a network. The server 4 includes a processing unit (CPU) 40, a storage unit 50, and an interface unit 60. Note that users in the present invention include not only owners of pet animals, but also agents, breeders, and the like. Pets include dogs, cats, rabbits, and birds, with dogs and cats being preferred.

Further, in the present embodiment, the processing calculation unit (CPU) 40 includes a calculation unit 41 and a prediction calculation unit 42, the storage unit 50 includes at least a setting unit 51, and the interface unit 60 includes an acquisition unit 61 and a prediction calculation unit 42. It is composed of an output section 62. The setting unit 51 may be configured to store mathematical formulas, functions, tables, or software related to the correlation between intestinal flora diversity data and beauty level. In the case of such a configuration, the calculation unit 41 calls a mathematical formula, function, table, or software stored in the setting unit 51 based on the basic information of the pets kept by the user, and calculates the variety of pets kept by the user. Beauty level can be calculated based on sex data. The beauty level refers to the level (level) of how good or bad a pet's appearance is, and includes, for example, the luster of the pet's fur, the quality of its coat, whether it has an appropriate weight, its body shape, and the luster of its skin. These include skin conditions such as mucus, eye mucus, and cloudy eyes. Since the beauty level is a level, it is preferable to output it in the form of a relative value when compared with an average value or a class when divided, rather than an absolute value. It is also possible to display numerical values in the form of deviation values. There are no particular restrictions on the specific output of the beauty level; for example, the level for each item such as hair gloss, coat texture, whether or not it is an appropriate weight, and body shape, or the level for each item may be summed or A format may also be used in which a total score is calculated by calculating an average value, and an evaluation value based on the total score is output. Output formats include methods that display relative numerical values such as deviation values as described above, methods that display classifications such as "good", "average", and "bad", "+1", "+2", and " Examples include a method of displaying a value when compared with an average value, such as "0" or "-1".

Here, the calculation unit 41 calculates the beauty level of the pet kept by the user based on the diversity data of the pet kept by the user and the correlation. The embodiment of FIG. 25 further includes a prediction calculation unit 42. The prediction calculation unit 42 predicts the future beauty level of the pet pet kept by the user based on the information on the food given to the pet. For example, the prediction calculation unit 42 predicts the future beauty level of the pet animal raised by the user. The beauty level is calculated by making corrections based on the information.
Further, the setting unit 51 sets the correlation between the pet animal's intestinal bacterial flora diversity data and the beauty level. As described above, formulas and functions related to correlations may be stored in advance, or formulas and functions may be newly set based on correlations.
Further, the acquisition unit 61 acquires diversity data of intestinal flora, preferably basic information of the pet kept by the user, and the output unit 62 outputs the beauty level etc. calculated by the calculation unit 41 to the user. Send.

[Beauty level estimation method]
FIG. 26 is a diagram illustrating an overview of a beauty level estimation method according to an embodiment of the present disclosure. As shown in FIG. 26, the beauty level estimation method according to the present embodiment includes a step S11 of acquiring basic information of a pet kept by a user and diversity data of intestinal flora, and Step S12 of setting a correlation between the diversity data of the flora and the beauty level, and calculating the beauty level of the pet pet kept by the user based on the diversity data of the pet pet kept by the user and the correlation. Step S13 is provided in this order.

Here, it is preferable that the method further includes a step of acquiring basic information about the pet kept by the user, and it is preferable to adjust the beauty level based on the basic information about the pet kept by the user. By taking into account the basic information of the pet pet kept by the user, it is possible to calculate a beauty level more suitable for the pet animal. For example, by adding the information that the age is 5 years old, the correlation shown in FIG. 28 can be used instead of that shown in FIG. 27, so that the fur gloss level of the pet can be estimated more accurately.

Furthermore, it may include step S14 of predicting the future beauty level of the pet pet kept by the user based on information about the food that the pet pet kept by the user eats. As explained in the insurance premium calculation system, the variety increases depending on the food that the pet eats, so by taking into account the food information, the future beauty level can be calculated.

[Aspects using trained models]
The beauty level estimation system of the present invention may employ a trained model. That is, a beauty level estimation system according to another embodiment of the present invention includes an acquisition unit that acquires diversity data of intestinal flora of a pet kept by a user, and a learned model. A beauty level estimation system comprising: a calculation unit that calculates a pet animal's beauty level from the acquired intestinal bacteria diversity data, wherein the trained model calculates the pet animal's beauty level from the intestinal bacteria diversity data. The model is characterized in that it is a learned model that has learned the relationship between this and the pet's beauty level.

A trained model can be generated, for example, by supervised learning or unsupervised learning. In the case of supervised learning, teacher data includes, for example, diversity data of an animal's intestinal flora and data and labels regarding the animal's beauty level.

Examples of methods for identifying the beauty level to be used as training data include methods to identify the beauty level based on questionnaires to owners, statements made when applying for pet insurance, questionnaires from experts such as groomers, etc. .

Artificial intelligence (AI) is preferable as the trained model. Artificial intelligence (AI) is software or systems that use computers to imitate the intellectual tasks performed by the human brain. A computer program that learns from experience. Artificial intelligence may be general-purpose or specialized, and specifically includes logistic regression, decision trees, k-means, multilayer perceptrons, recurrent neural networks, deep neural networks, convolutional neural networks, etc. Any publicly known software may be used, and any publicly available software may be used.

Train artificial intelligence to generate a trained model. The learning may be either machine learning or deep learning, but deep learning is preferable. Deep learning is a development of machine learning, and is characterized by automatically finding feature values.

The learning method for generating a trained model is not particularly limited, and publicly available software or libraries can be used. The learning method may be transfer learning. For example, DIGITS (the Deep Learning GPU Training System) published by NVIDIA can be used. Further, for example, a trained model can be generated by transfer learning using ResNet, MobileNet, or EfficientNet as an artificial intelligence (neural network) and Pytorch or the like as a machine learning library (deep learning library). In addition, for example, a known support vector machine published in "Introduction to Support Vector Machines" (Kyoritsu Shuppan) may be used.

The present inventors have found that there is a correlation between the diversity data of the intestinal flora of an animal and the beauty level of the animal, as shown in the reference example. Therefore, if learning is performed using the diversity data of an animal's intestinal flora and the beauty level of the animal as training data, it is possible to compare the diversity data of the intestinal flora of a pet and the beauty level of the pet. A trained model that has learned the relationship, particularly a trained model that uses the pet animal's intestinal flora diversity data as input and the output as the pet animal's beauty level, is obtained.

[Comprehensive health estimation system]
Furthermore, the present invention provides a comprehensive health level estimation system that calculates a comprehensive health level based on the beauty level and insurance risk. Here, the index regarding the overall health level can be set as appropriate by the user. For example, beauty level is an indicator for determining whether a person is not sick (not sick but not healthy), while insurance risk is an indicator that reflects a disease that has become manifest, so the weights of each are adjusted. It is also possible to specify that the values are added together. Specifically, the specification is such that the value obtained by subtracting the value obtained by multiplying the beauty level by 0.1 from the insurance risk is the index of the overall health level.

[Beauty level estimation method]
The beauty level estimating method of the present invention includes the step of acquiring the diversity data of the intestinal flora of a pet animal, and the computer calculating the diversity data of the intestinal flora of the pet animal and the beauty level. and calculating the beauty level of the pet kept by the user based on the correlation. Furthermore, it is preferable to include a step of acquiring basic information about a pet kept by the user, and it is preferable that the basic information about the pet is at least the breed and age.
Moreover, the beauty level estimation method of the present invention is based on the diversity data of the pet pet kept by the user, the correlation between the diversity data of intestinal bacteria of the pet animal, and the beauty level. Preferably, the method further includes a step of correcting the beauty level derived in the step of calculating the animal's beauty level based on information on food eaten by the pet animal kept by the user. The diversity data, beauty level, and correlation are the same as the beauty level estimation system described above.

Next, technical features of the present invention will be explained in detail based on reference examples.
[Reference example 5]
[Calculation of beauty level based on diversity data]
Hereinafter, an example of a method of calculating a beauty level based on diversity data will be described with reference to FIG. 27. FIGS. 27 to 29 are graphs showing the correlation between the diversity data of intestinal flora and the beauty level, which were collected and plotted by the present inventors.
Figure 27 shows the correlation between the coat gloss level, which is a type of beauty level, and the diversity index for 1-7 year old dogs (134,813 dogs) with no breed restrictions.The vertical axis is the coat gloss level, The horizontal axis is the diversity index (Shannon index). As shown in FIG. 27, it can be seen that there is a correlation between data regarding diversity and coat gloss level.

Here, a method for setting the hair gloss level will be explained. First, in order to perform statistical analysis, the diversity index was divided into four categories: (1) 2.0 or more, less than 3.0, (2) 3.0 or more, less than 4.0, (3) 4 (4) 5.0 or more, less than 6.0). For each category, a person with specialized knowledge (a trimmer) judged the hair gloss level based on the gloss and shine of the hair. More specifically, among dogs aged 1-7 years with no breed restrictions, the average value of coat luster for a group of dogs with a diversity index of 2.0 or more but less than 3.0 is ``0 (intermediate standard)''. The average value of coat luster for groups with an index of 5.0 or more and less than 6.0 is "1 (good standard)", and among 7-year-old dogs with no breed restrictions, the diversity index is 2.0 or more and less than 3.0. The average value of the coat gloss of the group was evaluated as "-1 (bad standard)".

Figure 28 shows the correlation between coat gloss level and diversity index for 4- to 7-year-old dogs (48,563 dogs) with no breed restrictions, and for comparison, 1- to 7-year-old dogs with no breed restrictions. Data is also included. As shown in FIG. 28, it can be seen that a correlation can be confirmed even if the age classification is changed, and that in the older age group, there is a large change in coat gloss due to a change in the diversity index.

Moreover, the beauty level is not limited to just the luster of the hair. For example, Figure 29 shows the correlation between standard body weight (the range of neither thin nor fat) and the diversity index, where the vertical axis is the proportion (%) of individuals with normal weight, and the horizontal axis is the diversity index. index (Shannon index). As shown in FIG. 29, a correlation is established not only with hair luster but also with standard body weight, indicating that standard body weight can be used as an index of beauty level.
Note that, in the present invention, an index that combines standard body weight and coat gloss may be set.

Claims (12)

an acquisition unit that acquires diversity data of the intestinal flora of pets kept by the user;
a calculation unit that calculates the beauty level of a pet pet kept by the user based on the intestinal flora diversity data and the correlation between the intestinal flora diversity data of the pet animal and the beauty level;
A beauty level estimation system equipped with. 2. The beauty level estimation system according to claim 1, wherein the acquisition unit further acquires basic information of a pet kept by the user. 3. The beauty level estimation system according to claim 1, wherein the beauty level is a level related to hair luster or a level related to body shape. 2. The beauty level estimation system according to claim 1, wherein the diversity data of the pet's intestinal flora is a Shannon index. The calculation unit corrects the beauty level calculated based on the basic information of the pet according to the beauty variation factor derived using the diversity data of the intestinal flora of the pet, thereby calculating the beauty level. 3. The beauty level estimation system according to claim 2, wherein the beauty level estimation system calculates . 3. The beauty level estimation system according to claim 2, wherein the basic information of the pet is at least breed and age. An acquisition unit that acquires the diversity data of the intestinal flora of a pet kept by the user, and a learned model are used to determine the beauty level of the pet from the diversity data of the intestinal bacteria acquired by the acquisition unit. A beauty level estimation system comprising: a calculation unit for calculating,
A beauty level estimation system characterized in that the trained model is a trained model that has learned a relationship between diversity data of a pet animal's intestinal flora and a beauty level of the pet animal. an acquisition unit that acquires diversity data of the intestinal flora of pets kept by the user;
a calculation unit that calculates the beauty level of a pet pet kept by the user based on the intestinal flora diversity data and the correlation between the intestinal flora diversity data of the pet animal and the beauty level;
a risk calculation unit that calculates the insurance risk of the pet pet kept by the user based on the intestinal flora diversity data and the correlation between the intestinal flora diversity data of the pet animal and the insurance risk; ,
A health level calculation section that calculates your health level based on your beauty level and insurance risk;
Comprehensive health level estimation system. obtaining diversity data of the pet's intestinal flora;
a step in which the computer calculates the beauty level of the pet pet kept by the user based on the diversity data and the correlation between the diversity data of the intestinal flora of the pet animal and the beauty level;
A beauty level estimation method that includes the following in order. 10. The beauty level estimation method according to claim 9, further comprising the step of acquiring basic information of a pet kept by the user. 11. The beauty level estimation method according to claim 10, wherein the basic information of the pet is at least breed and age. derived in the step of calculating the beauty level of the pet kept by the user based on the diversity data of the pet kept by the user and the correlation between the diversity data of intestinal bacteria of the pet and the beauty level. The beauty level estimating method according to claim 10 or 11, further comprising the step of correcting the beauty level based on information about food eaten by a pet kept by the user.