Disease predictiion system, insurance premium calculation system, and disease prediction method
The disease prediction system uses insurance data and animal health history to predict neoplastic diseases, addressing the challenge of early detection and enabling proactive prevention, thus reducing medical costs and improving animal health outcomes.
Patent Information
- Application Number
- PCT/JP2026/016743
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-07-04
- Filing Date
- 2026-04-27
- Publication Date
- 2026-11-12
AI Technical Summary
Early detection of neoplastic diseases in animals is difficult due to the lack of widespread regular health checkups, and existing methods do not provide a simple way to predict the likelihood of such diseases, leading to increased medical expenses and mortality.
A disease prediction system that utilizes insurance claim data to identify correlations between previous diseases and the incidence of neoplastic diseases, incorporating factors like benign tumors, gut microbiota diversity, and animal health history to predict the likelihood of neoplastic diseases using a prediction unit and insurance premium calculation system.
Enables early prediction of neoplastic diseases, allowing for proactive preventive measures and potentially reducing the risk of malignant tumors by identifying weakened immunity, thereby reducing medical expenses and improving animal health management.
Smart Images

Figure JP2026016743_12112026_PF_FP_ABST
Abstract
Description
Disease Prediction System, Insurance Premium Calculation System, and Disease Prediction Method
[0001] The present invention relates to a disease prediction system, an insurance premium calculation system, and a disease prediction method, and more specifically, to a disease prediction system, an insurance premium calculation system, and a disease prediction method that provide information on the possibility of an animal contracting a neoplastic disease in the future from data on the animal's history of contracting specific diseases.
[0002] Companion animals such as dogs, cats and rabbits, and livestock such as cattle and pigs are irreplaceable existences for humans. In recent years, while the average lifespan of animals raised by humans has increased significantly, animals are more likely to contract some kind of disease during their lifetime, and the increase in medical expenses borne by breeders has become a problem.
[0003] In order to maintain the health of animals, physical condition management through daily diet, exercise, etc. and prompt response to discomfort are important. However, since animals cannot express their physical discomfort in their own words, the actual situation is that breeders only notice that the animal has contracted a disease when symptoms progress and some externally observable signs appear.
[0004] In particular, neoplastic diseases (cancers) are serious diseases that rank high in mortality not only in humans but also in animals. However, the current situation is that early detection is difficult for animals where regular health checkups are not widespread.
[0005] Therefore, there is a demand for a means to know, in a simple manner, whether an animal is likely to contract a neoplastic disease in the future. In particular, if it is possible to know the possibility of contracting a neoplastic disease in the future in a state where the animal has not contracted a neoplastic disease and no symptoms specific to the neoplastic disease have appeared, it is useful because specific measures for the prevention of neoplastic diseases can be taken.
[0006] Patent Document 1 discloses a gut microbiota regulating or improving composition that has the effect of effectively regulating or improving the gut microbiota by proliferating bacteria of the phylum Bacteroidetes and reducing bacteria of the phylum Firmicutes in the gut microbiota, but does not disclose a method for predicting animal diseases.
[0007] International Publication 2017 / 094892 Brochure
[0008] Therefore, the present invention aims to provide a disease prediction system, etc., that can predict the likelihood of an animal developing a neoplastic disease in the future using a simple method.
[0009] The inventors examined insurance claim data for tens of thousands of animals accumulated by pet insurance companies and investigated whether there were subsequent insurance claims for neoplastic diseases between groups of animals that had filed insurance claims for a specific disease within a predetermined period (affected group) and groups that had not filed insurance claims for the same disease (unaffected group). They found that there was a difference in subsequent insurance claims for neoplastic diseases, i.e., the prevalence of neoplastic diseases, between the affected group and the unaffected group, and that it is possible to predict neoplastic diseases from information regarding the presence or absence of a certain disease, thus completing the present invention.
[0010] Furthermore, the inventors, after analyzing a vast number of insurance claim records, found a statistical correlation between the prevalence of benign tumors and the incidence of malignant tumors within a predetermined period, such as one year, from the time of an insurance claim due to a benign tumor. This suggests that benign tumors may be a sign of malignant tumors. Traditionally, benign tumors have been considered clinically harmless because the immune system tolerates their existence (immune tolerance). The inventors' findings could overturn this common understanding in the veterinary medical community. The inventors consider the following to be one mechanism by which a statistical correlation is found between the prevalence of benign tumors and the incidence of malignant tumors, but the present invention is not limited to this mechanism. That is, benign tumors include not only typical benign tumors that are well monitored by the immune system and are considered clinically harmless, but also benign tumors that develop after escaping immune surveillance due to a situation where acquired immunity is insufficient, such as a decrease in the diversity of the gut microbiota or inflammation caused by periodontal disease-related bacteria. As a result of such a continued state of immune deficiency, it is possible that the initially discovered benign tumor will not become malignant, but rather another malignant tumor will develop. With this invention, when a benign tumor is discovered in an animal, instead of simply observing it as in the past, it will be possible to recognize the benign tumor as a sign of weakened immunity in the animal or a sign of increased risk of developing a malignant tumor, and to take measures to suppress the development of malignant tumors, such as taking measures to strengthen immunity. This is expected to establish a new preventive medicine approach in veterinary medicine.
[0011] In other words, the present invention is as follows: [1] to
[16] [1] A disease prediction system characterized by comprising a prediction unit that predicts whether an animal will contract a neoplastic disease, using information on the incidence of a predetermined disease excluding neoplastic diseases in animals other than humans. [2] The disease prediction system according to [1], wherein the prediction unit evaluates the probability of contracting a neoplastic disease to be low if the animal other than humans has previously suffered from at least one of valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes. [3] A disease prediction system characterized by comprising a correlation determination unit that compares the incidence rate of neoplastic diseases in an animal group that has suffered from a predetermined disease excluding neoplastic diseases with the incidence rate of neoplastic diseases in an animal group that has never suffered from a predetermined disease excluding neoplastic diseases, and determines that the predetermined disease is correlated with neoplastic diseases if the incidence rates of neoplastic diseases differ between the animal groups. [4] An insurance premium calculation system characterized by comprising an insurance premium calculation unit that calculates the insurance premium for an animal using information on the incidence of a predetermined disease excluding neoplastic diseases in animals other than humans. [5] A disease prediction method characterized by comprising a prediction step in which a computer uses information on the incidence of predetermined diseases other than neoplastic diseases in animals other than humans to predict whether the animal will contract a neoplastic disease. [6] A premium calculation method characterized by comprising a premium calculation step in which a computer uses information on the incidence of predetermined diseases other than neoplastic diseases in animals other than humans to calculate the premium for the animal. [7] The disease prediction system of [1] in which the prediction unit highly evaluates the possibility of contracting a neoplastic disease if the animal other than humans has previously suffered from at least one of vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis. [8] A disease prediction system characterized by comprising a prediction unit that uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will contract a neoplastic disease. [9] The disease prediction system of [8] in which the prediction unit highly evaluates the possibility of contracting a neoplastic disease if the animal other than humans has previously suffered from a benign tumor.
[10] The disease prediction system of [8], wherein the prediction unit further uses information about the animal's weight to predict whether the animal will suffer from a neoplastic disease.
[11] The disease prediction system of [8], wherein the prediction unit further uses information on the diversity of the animal's intestinal microbiota to predict whether the animal will contract a neoplastic disease.
[12] A mortality prediction system characterized by comprising a mortality prediction unit that uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will die within a predetermined period.
[13] A disease prediction system characterized by comprising a kidney disease prediction unit that uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will contract chronic kidney disease.
[14] An insurance premium calculation system characterized by comprising an insurance premium calculation unit that uses information on the incidence of benign tumors in animals other than humans to calculate the insurance premium for the animal.
[15] A disease prediction method characterized by comprising a prediction step in which a computer uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will contract a neoplastic disease.
[16] An insurance premium calculation method characterized by comprising an insurance premium calculation step in which a computer uses information on the incidence of benign tumors in animals other than humans to calculate the insurance premium for the animal.
[0012] The present invention makes it possible to provide a disease prediction system, etc., that can predict the likelihood of an animal developing a neoplastic disease in the future using a simple method.
[0013] This is a schematic diagram of the disease prediction system of the present invention. This is a flowchart of disease prediction using the disease prediction system of the present invention. This is a graph showing the relationship between the history of vomiting, diarrhea, and bloody stool diseases in dogs and the incidence (prevalence) of neoplastic diseases. This is a graph showing the relationship between the history of atopic dermatitis in dogs and the incidence (prevalence) of neoplastic diseases. This is a graph showing the relationship between the history of allergic dermatitis in dogs and the incidence (prevalence) of neoplastic diseases. This is a graph showing the relationship between the history of skin tumors (benign tumors) in dogs and the incidence (prevalence) of neoplastic diseases. This is a graph showing the relationship between the history of valvular heart disease in dogs and the incidence (prevalence) of neoplastic diseases. This is a graph showing the relationship between the history of epilepsy in dogs and the incidence (prevalence) of neoplastic diseases. This is a graph showing the relationship between the history of benign tumors in dogs and the incidence (prevalence) of neoplastic diseases when dogs are divided into mixed breeds and purebreds. This graph shows the relationship between a dog's history of benign tumors and its mortality rate in the following year. This graph shows the relationship between age and the incidence (prevalence) of benign tumors, separated into mixed breeds and purebreds. This graph shows the relationship between a dog's weight and the incidence (prevalence) of benign tumors. This graph shows the relationship between the weight of mixed breed dogs and the incidence (prevalence) of benign tumors. This graph shows the relationship between the weight of poodles and the incidence (prevalence) of benign tumors. This graph shows the relationship between the diversity of a dog's gut microbiota and the incidence (prevalence) of benign tumors. This graph shows the relationship between the diversity of a dog's gut microbiota and the incidence (prevalence) of lipomas. This graph shows the relationship between the diversity of a dog's gut microbiota and the incidence (prevalence) of histiocytomas. This graph shows the relationship between a dog's weight and the incidence (prevalence) of neoplastic diseases. This graph shows the relationship between the diversity of a dog's gut microbiota and the incidence (prevalence) of neoplastic diseases. This graph shows the relationship between the diversity of the gut microbiota in dogs and the incidence (prevalence) of lymphoma. It also includes an example of a risk assessment score for a dog's history of benign tumors and gut microbiota diversity. Furthermore, it shows the relationship between a risk assessment score for a dog's history of benign tumors and gut microbiota diversity and the incidence (prevalence) of neoplastic diseases. Finally, it presents an example of a risk assessment score for a cat's history of benign tumors and gut microbiota diversity.This graph shows the relationship between a cat's history of benign tumors, the diversity of its gut microbiota, and the incidence (prevalence) of chronic kidney disease. It also shows the relationship between a risk assessment score assigned to a cat's history of benign tumors and the diversity of its gut microbiota, and the incidence (prevalence) of chronic kidney disease. Furthermore, it illustrates the relationship between a cat's history of benign tumors and the incidence of neoplastic diseases. Finally, it shows the relationship between a cat's body weight and the incidence of neoplastic diseases. It also illustrates the relationship between a cat's body weight and the incidence of benign tumors. The graph also shows the relationship between a cat's body weight and the incidence of skin tumors. And finally, it illustrates the relationship between a cat's body weight and the incidence of non-skin neoplastic diseases.
[0014] [Disease Prediction System] The disease prediction system of the present invention is characterized by comprising a prediction unit that predicts whether an animal other than a human will contract a neoplastic disease, using information on the incidence of diseases such as valvular heart disease, epilepsy, Malassezia otitis externa, diabetes mellitus, vomiting, diarrhea, bloody stools, allergic dermatitis, atopic dermatitis, or periodontal disease in the animal. Hereinafter, diseases other than neoplastic diseases in animals other than humans, such as valvular heart disease (mitral valve insufficiency), epilepsy, Malassezia otitis externa, diabetes mellitus, vomiting, diarrhea, bloody stools, allergic dermatitis, atopic dermatitis, and periodontal disease, will be referred to as "specified diseases." In the case of "neoplastic diseases," benign tumors are excluded.
[0015] In other words, the specific diseases in the disease prediction system of the present invention are not particularly limited and include, for example, skin diseases, ear diseases, musculoskeletal diseases, ophthalmic diseases, digestive diseases, systemic diseases, urinary tract diseases, liver, biliary tract and pancreatic diseases, circulatory diseases, nervous system diseases, respiratory diseases, dental and oral diseases, endocrine diseases, reproductive system diseases, and blood and hematopoietic system diseases. Examples of skin diseases include dermatitis, atopic dermatitis, and pyoderma. Examples of ear diseases include otitis externa and otitis media. Examples of musculoskeletal diseases include patellar luxation and herniated discs. Examples of ophthalmic diseases include conjunctivitis, eye discharge, keratitis, corneal ulcer / erosion, epiphora, cataracts, and glaucoma. Examples of digestive diseases include gastritis and enteritis. Examples of systemic diseases include lethargy and collapse. Examples of urinary tract diseases include cystitis and urolithiasis. Examples of hepatobiliary and pancreatic diseases include biliary sludge and chronic renal failure. Examples of circulatory system diseases include valvular heart disease and cardiomyopathy. Examples of nervous system diseases include epilepsy and seizures. Examples of respiratory system diseases include coughing, rhinitis, tracheal collapse, and bronchial stenosis. Examples of dental and oral system diseases include periodontal disease and stomatitis. Examples of endocrine system diseases include hypothyroidism and diabetes mellitus. Examples of reproductive system diseases include mammary gland tumors and balanitis. Examples of blood and hematopoietic system diseases include lymphoid tumors and thrombocytopenia. Furthermore, the specific disease in the disease prediction system of the present invention may be any one of these diseases, or it may be a general term that includes other diseases. Furthermore, the disease in the present invention is not limited to any particular disease, and may include symptoms such as lethargy and tear staining.
[0016] Neoplastic diseases are not particularly limited and include, for example, various cancers such as stomach cancer, lung cancer, esophageal cancer, skin cancer, oral cancer, and tongue cancer, as well as cell tumors, mesothelioma, sarcoma, brain tumors, lymphoma, neuroblastoma, and leukemia. Note that when referring to "neoplastic diseases," benign tumors are excluded. Benign tumors are tumors that do not show malignant findings pathologically, and include, for example, papillomas, trichoblastomas, sebaceous gland hyperplasia / epitheliomas, lipomas, histiocytomas such as cutaneous histiocytomas, mast cell tumors, and perianal tumors. Lipomas and histiocytomas are preferred. Whether a tumor is benign can be determined by known methods, such as by taking and observing cells through needle biopsy or excisional biopsy. The determination of whether a tumor is benign is preferably made by a veterinarian.
[0017] Another embodiment of the present invention is a disease prediction system that includes a prediction unit that uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will develop a neoplastic disease. Hereinafter, a disease prediction system that uses information on the incidence of a specific disease to predict whether an animal will develop a neoplastic disease and a disease prediction system that uses information on the incidence of benign tumors to predict whether an animal will develop a neoplastic disease will be described together.
[0018] [Prediction Unit] Preferably, the prediction unit of the present invention uses a processor (arithmetic unit) such as a CPU or GPU in a computer or server to predict whether the target animal will be affected by a neoplastic disease, using a pre-set program, code, or software. The method of determination and prediction is not particularly limited, but as a preferred example, the processor uses information and data regarding the target animal's susceptibility to one of the specific diseases or benign tumors to predict whether it will be affected by a neoplastic disease.
[0019] In this invention, animals other than humans are preferably mammals, and dogs and cats are particularly preferred.
[0020] In this invention, "prediction" refers to a prediction of whether a target animal will develop a neoplastic disease within a predetermined period, for example, within 5 years, 3 years, or 1 year from the time the prediction is made. The form of the prediction is not particularly limited and may include, for example, predicting whether or not an animal will develop a neoplastic disease, indicating the probability of developing one, or determining the magnitude of the risk. When indicating the probability of developing a neoplastic disease, it is preferable to express it as a percentage between 0 and 100%. When determining the magnitude of the risk, it is preferable to express the magnitude of the risk in stages, for example, large (high), medium, or small (low).
[0021] The prediction unit preferably has a function to evaluate the likelihood of developing a neoplastic disease based on whether or not an animal other than a human has ever suffered from a specific disease or a benign tumor. For example, if the animal has suffered from a specific disease, particularly vomiting, diarrhea, and bloody stool disease, allergic dermatitis, atopic dermatitis, or a benign tumor, or if it has suffered from the above specific disease or a benign tumor within a predetermined period preceding the prediction time, the likelihood of developing a neoplastic disease may be evaluated as high. Evaluating the likelihood of developing a neoplastic disease as high means evaluating the probability of developing the disease as high or overestimating the risk of developing the disease. On the other hand, if the animal has never suffered from a specific disease, particularly vomiting, diarrhea, and bloody stool disease, allergic dermatitis, atopic dermatitis, or a benign tumor, it is preferable to evaluate the likelihood of developing a neoplastic disease as low. Evaluating the likelihood of developing a neoplastic disease as low means evaluating the probability of developing the disease as low or underestimating the risk of developing the disease.
[0022] Furthermore, the prediction unit may have a function to evaluate the likelihood of developing a neoplastic disease depending on the disease. For example, the prediction unit can evaluate the likelihood of developing a neoplastic disease as high if an animal other than a human has a history of vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis. Alternatively, the prediction unit may evaluate the likelihood of developing a neoplastic disease as low if an animal has a history of valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes mellitus.
[0023] The prediction may use information other than information regarding the incidence of any specific disease in animals other than humans. Such information may include physical data such as the animal's age, breed, sex, weight, height, and length, as well as sequence data such as genome sequences, SNPs, and specific gene sequences, medical history, family history, and information regarding the diversity of the gut microbiota. Preferably, information regarding weight is used for prediction. The inventors have found a correlation between the weight of an animal and the incidence of neoplastic diseases. By utilizing this correlation, it is expected that the accuracy of predicting whether or not an animal will develop a neoplastic disease can be improved by using information regarding the weight of the animal being predicted. It is preferable to predict that the higher the weight of the animal being predicted, the greater the risk of developing a neoplastic disease. As information regarding weight, the numerical value of the animal's weight may be used, or a classification that divides the weight into predetermined ranges may be used. A classification that divides the weight into predetermined ranges may, for example, be divided into large, medium, small, and very small according to weight, and the information of which classification the animal belongs to is used for prediction. It is also preferable to use information regarding the diversity of the gut microbiota. Information on the diversity of the gut microbiota includes the Shannon Index (Shannon-Wiener diversity index) and the Simpson Diversity Index. Information on the diversity of the gut microbiota can be obtained using known metagenomic analysis methods such as amplicon sequencing using sequencers such as NGS, or microbiota analysis methods, using samples such as feces collected from the target animals.
[0024] Information regarding the occurrence of any of the specified diseases or benign tumors in animals other than humans refers to information on whether or not an animal has ever been affected by any of the specified diseases or benign tumors, preferably information on whether or not the animal was affected by any of the specified diseases or benign tumors before the prediction by the prediction unit was performed. More preferably, it refers to information on whether or not the animal was affected by any of the specified diseases or benign tumors within 10 years, 5 years, 3 years, or 1 year prior to the prediction by the prediction unit. The form and input method of the information are not particularly limited, and known forms and input methods can be adopted.
[0025] The method for obtaining information on the incidence of any of the specified diseases or benign tumors in animals other than humans is not particularly limited. For example, it can be obtained and entered through pet insurance claim records, input by users of the system of the present invention, declarations, uploads, etc. For example, one method is to upload information on the incidence of any of the specified diseases or benign tumors, along with basic information such as the age, sex, species, and breed of the animal to be predicted, from a terminal used by the user via an internet connection to the system of the present invention.
[0026] Furthermore, the prediction unit may perform predictions and judgments using a pre-trained model. Examples of pre-trained models include a model that has learned the relationship between information on the presence or absence of a specific disease or benign tumor and the presence or absence of a neoplastic disease. Such a pre-trained model can be obtained, for example, by training it with information on whether an animal has been affected by one of the specific diseases or a benign tumor within a predetermined period, and information on whether that animal subsequently developed a neoplastic disease, as training data. Whether or not an animal has been affected by a disease can be replaced with a dummy variable. Information on whether or not an animal has been affected by a disease can be obtained, for example, from a veterinary hospital or the insured owner as an insurance claim (also called an "accident"). In other words, if the animal is insured as a pet, and the animal is diagnosed with a disease after visiting a hospital, the veterinary hospital or the owner (the pet insurance policyholder) will claim insurance payment from the insurance company along with the fact of the disease, so the insurance company can learn that the animal has been affected by a disease. On the other hand, if no claim for insurance benefits is made within a specified period of time, it can be determined that the insured animal did not suffer from the disease during that period.
[0027] The aforementioned trained model is preferably artificial intelligence (AI). Artificial intelligence (AI) is software or a system that imitates the intellectual work performed by the human brain, and specifically refers to computer programs that understand natural language used by humans, perform logical reasoning, and learn from experience. The AI can be general-purpose or specialized, and can be any type of network, such as a deep neural network or a convolutional neural network, and publicly available software can be used.
[0028] To generate a trained model, artificial intelligence is trained using training data. Either machine learning or deep learning can be used for training, but machine learning is preferred. Deep learning is an advanced form of machine learning and is characterized by its ability to automatically identify features.
[0029] There are no particular restrictions on the training method used to generate a pre-trained model, and publicly available software can be used. For example, NVIDIA's DIGITS (the Deep Learning GPU Training System) can be used. Alternatively, the model may be trained using a publicly known support vector machine method, such as the one published in "Introduction to Support Vector Machines" (Kyoritsu Shuppan).
[0030] Machine learning can be either unsupervised or supervised, but supervised learning is preferred. Supervised learning methods are not limited to any particular type; examples include decision trees, ensemble learning, and gradient boosting. Publicly available machine learning algorithms include, for example, XGBoost, CatBoost, and LightGBM.
[0031] The trained models may be generated for each individual disease, such as benign tumors or specific diseases, or they may be generated for multiple diseases at once.
[0032] [Reception Unit] The disease prediction system of the present invention preferably includes a reception unit. The reception unit is configured to receive input of information regarding the occurrence of any of the specified diseases or benign tumors in animals other than humans. The method for receiving such information can employ known methods and configurations, and may include, for example, inputting data into a terminal, transmitting data from the terminal to a server via a network, or uploading. Examples of the configuration of the reception unit include internal / external interfaces such as SCSI and SAS, and network-specific interfaces such as FC (Fibre Channel), iSCSI, FCoE, and NAS. Other examples of the configuration of the reception unit include touch panels, buttons, and keys for inputting information.
[0033] [Output] When the prediction unit of the present invention receives information or data as input information regarding the presence of one of the specified diseases or a benign tumor, the processor uses a program to predict whether the animal will develop a neoplastic disease, or whether it will develop a disease within a predetermined period, preferably within three years, more preferably within one year, from a certain point in time, such as when information about the presence of one of the specified diseases is obtained. The output format is not particularly limited, and for example, the prediction can be output by displaying on the screen of a terminal such as a personal computer, "There is a high possibility that the animal has a neoplastic disease," "There is a possibility that the animal will develop a neoplastic disease within the next year," or "There is a low possibility that the animal will develop a neoplastic disease within the next year." The disease prediction system of the present invention may also have a separate output unit that receives the judgment result from the prediction unit and outputs the judgment result. Examples of the output unit include an interface involved in the exchange of information with the outside, and a display that outputs the result.
[0034] Hereinafter, one embodiment of the disease prediction system of the present invention will be described with reference to Figure 1.
[0035] In Figure 1, terminal 40 is a terminal used by the user. Examples of terminal 40 include personal computers and tablet terminals. Terminal 40 is composed of a processing unit such as a CPU, a storage unit such as a hard disk, ROM or RAM, a display unit such as an LCD panel, an input unit such as a mouse, keyboard, or touch panel, and a communication unit such as a network adapter. The user accesses the server from terminal 40 and inputs and transmits information about the medical history of the animal to be predicted, a facial image (photograph), and information such as the type, breed, age at the time of photography, and weight of the animal. The user can also receive disease prediction results from the server by having terminal 40 access the server.
[0036] In this embodiment, the server is configured as a computer, but any device may be used as long as it has the functions according to the present invention. The storage unit 10 is composed of, for example, ROM, RAM, or a hard disk. The storage unit 10 stores information processing programs for operating each part of the server, and in particular stores the prediction unit 11 and, if configured as an insurance premium calculation system, the insurance premium calculation unit 12. If the purpose is not to calculate insurance premiums, but simply to output a prediction of the incidence of neoplastic diseases, the insurance premium calculation unit 12 may be omitted.
[0037] As described above, the prediction unit 11 takes as input information about the medical history of the animal to be predicted, entered by the user, and if the medical history includes a specific disease or a benign tumor, it uses information about the animal's susceptibility to either the specific disease or the benign tumor to output a prediction of whether or not the animal will develop a neoplastic disease within a predetermined period (for example, within six months) from the time the information is entered.
[0038] The insurance premium calculation unit 12 is software that calculates the insurance premium for an animal based on the prediction of the incidence of neoplastic disease output by the prediction unit 11 and information such as the type, breed, age at the time of acquisition of the intestinal microbiota data, and weight of the animal, which is entered by the policyholder. For example, the software classifies the insurance premium into grades according to the type, breed, age at the time of acquisition of the intestinal microbiota data, etc., and finally adjusts the grade by taking into account the prediction of the incidence of disease output by the prediction unit 11, and calculates the final insurance premium. The insurance premium calculation unit 12 and the prediction unit 11 may be configured to use a single software.
[0039] The processing unit 20 is composed of processors such as a CPU and a GPU, and uses programs stored in the memory unit for the prediction unit 11 and the insurance premium calculation unit 12 to predict the incidence of neoplastic diseases and calculate insurance premiums.
[0040] The interface unit (communication unit) 30 includes a reception unit 31 and an output unit 32, and receives data on the animal's medical history, information on whether or not it has any of the specified diseases or benign tumors, and other information from the user's terminal, and outputs predictions of the incidence of neoplastic diseases and calculation results of insurance premiums to the user's terminal.
[0041] With the disease prediction system of this embodiment, users can input information about whether or not an animal, such as a pet, has a specific disease or benign tumor, and obtain a prediction result indicating whether or not the animal is at risk of developing a neoplastic disease, or whether the risk is high or low.
[0042] FIG. 2 shows a flowchart of disease prediction and determination based on one embodiment of the disease prediction method using the disease prediction system of the present invention. A user accesses the disease prediction system via a terminal, and inputs data related to the presence or absence of affection with any of the specific diseases (or benign tumors) of an animal (Step S1). The disease prediction system predicts and determines, from the input data related to the presence or absence of affection with the specific disease (or benign tumor) of the animal, how high the possibility is that the animal will suffer from a neoplastic disease within a predetermined period (for example, within 1 year) or currently suffers from a neoplastic disease (Step S2). The disease prediction system outputs the prediction result, transmits it to the terminal 40, and the prediction result is displayed on the terminal 40 (Step S3).
[0043] [Embodiment Relating to Input of Disease Affection Information] Hereinafter, one embodiment relating to the input screen for disease affection information, prediction processing, and output of prediction results in the disease prediction system of the present invention will be described.
[0044] In the disease prediction system of the present embodiment, the receiving unit 31 causes the display unit of the terminal 40 used by the user to display an input screen for inputting whether an animal other than a human is suffering from valvular disease, epilepsy, Malassezia otitis externa, diabetes, vomiting / diarrhea / bloody stool, allergic dermatitis or atopic dermatitis, or has suffered from any of these within the past 3 years.
[0045] On the input screen, the aforementioned disease names are displayed as a list, and input elements for inputting the current affection status and past affection history are provided corresponding to each disease name. The input elements can be configured by known GUI (Graphical User Interface) elements such as check boxes, toggle switches, radio buttons, or pull-down menus.
[0046] Specifically, on the input screen, for example, respective disease names including "valvular disease", "epilepsy", "Malassezia otitis externa", "diabetes", "vomiting / diarrhea / bloody stool", "allergic dermatitis" and "atopic dermatitis" are displayed. For each disease name, a checkbox or toggle switch that allows selection of at least one of "currently suffering from the disease" and "have suffered from the disease in the past 3 years" is arranged. The user can input information related to morbidity by selecting (checking on) the corresponding item according to the morbidity status of the target animal.
[0047] The input screen may be configured to allow input of only "currently suffering from the disease" or only "have suffered from the disease in the past 3 years", or may be configured to allow input of both "currently suffering from the disease" and "have suffered from the disease in the past 3 years" at the same time. When configured to allow input of both, two checkboxes or toggle switches may be arranged side by side for each disease name, or it may be configured as a three-stage toggle that can distinguish between "currently suffering from the disease" and "have a past history of suffering from the disease" (for example, "none", "suffered in the past", "currently suffering from the disease").
[0048] In addition, disease names other than the above-mentioned respective diseases may be additionally displayed on the input screen, and a free text input field may be provided to allow the user to input any arbitrary disease name. It is preferable that the input screen is further provided with fields for inputting basic information such as the type of the target animal (dog, cat, etc.), breed, age, sex, and body weight.
[0049] [Embodiment of Prediction Processing] When the user inputs facts about current morbidity or past morbidity of the above-mentioned diseases on the input screen and transmits the input content, the receiving unit 31 receives the relevant input data. The prediction unit 11 uses the received relevant input data to make a prediction about the possibility of the target animal suffering from a neoplastic disease.
[0050] Specifically, the processors, such as the CPU and GPU that make up the processing unit 20, read and execute prediction software (prediction programs) that are pre-stored in the storage devices, such as the hard disk, ROM, or RAM that make up the storage unit 10. This allows the processors to function as a prediction unit 11 and perform predictions regarding the likelihood of developing neoplastic diseases.
[0051] Based on the input disease prevalence information, the prediction unit 11 evaluates the likelihood of developing a neoplastic disease in different directions, depending on the type of disease.
[0052] Firstly, the prediction unit 11 evaluates the likelihood of developing a neoplastic disease as low if the fact of having or having had valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes mellitus is entered. Specifically, for example, the prediction unit 11 calculates the prediction probability by multiplying the base likelihood of developing a neoplastic disease (reference likelihood), which is calculated when an animal does not have any of these diseases and has never had any of these diseases from the past to the present, by a negative coefficient or a positive coefficient less than 1 (for example, a coefficient in the range of 0.5 to 0.9). As a result, animals with a history of these diseases are predicted to have a lower likelihood of developing a neoplastic disease compared to animals without such a history.
[0053] Here, a "negative coefficient" may be a coefficient that represents a subtraction from the baseline probability of illness. For example, it may be a coefficient that reduces the predicted probability by adding (i.e., subtracting) a predetermined negative value to the baseline probability of illness. A "coefficient less than 1" may be a coefficient that reduces the predicted probability by multiplying the baseline probability of illness by a positive value less than 1. It is preferable to set the specific values of the coefficients based on statistical data such as past insurance claim data.
[0054] If a person has or has previously had multiple diseases from among valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes, they may apply a coefficient corresponding to each disease individually, or they may use a comprehensive coefficient based on the number of diseases they have.
[0055] Furthermore, the prediction unit 11 can differentiate the degree of influence on the likelihood of developing a neoplastic disease depending on whether the input indicates that the target animal is not currently affected but has been affected in the past, or whether the input indicates that the animal is currently affected. Specifically, it is preferable to evaluate the influence on the likelihood of developing a neoplastic disease more highly (i.e., predict a lower likelihood of developing a neoplastic disease) when the animal is currently affected, compared to when the animal has been affected in the past but is not currently affected.
[0056] For example, by setting a coefficient of 0.6 for animals currently suffering from valvular heart disease and a coefficient of 0.8 for animals that have had valvular heart disease in the past but are not currently affected, it is possible to predict a stronger reduction in the likelihood of developing neoplastic disease in animals currently suffering from the disease. This is because animals currently undergoing treatment are continuously receiving treatment and dietary management for the disease, and it is possible that the improvement in immune function is having a stronger effect.
[0057] Secondly, the prediction unit 11 evaluates the likelihood of developing a neoplastic disease highly if the fact of having or having had vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis is entered. Specifically, for example, the prediction unit 11 calculates the prediction probability by multiplying the standard likelihood of developing the disease by a positive coefficient or a coefficient greater than 1 (for example, a coefficient in the range of 1.1 to 1.5). As a result, animals with a history of these diseases are predicted to have a higher likelihood of developing a neoplastic disease compared to animals without such a history.
[0058] The reason for this is that animals suffering from vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis often have abnormal immune function or are in a chronic inflammatory state, and such conditions may increase the risk of developing neoplastic diseases.
[0059] Furthermore, the degree to which vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis influences the likelihood of developing neoplastic diseases may differ depending on whether the individual is currently suffering from these conditions or has suffered from them in the past but is not currently affected.
[0060] Furthermore, if the user has or has previously had both valvular heart disease and vomiting, diarrhea, bloody stools, etc., the prediction unit 11 calculates an overall prediction probability by combining coefficients corresponding to each disease. For example, it comprehensively evaluates the likelihood of developing the disease reduced by having valvular heart disease and the likelihood of developing the disease increased by having vomiting, diarrhea, bloody stools, etc., and calculates the final prediction probability.
[0061] [Embodiment concerning output of prediction results] When the prediction processing by the prediction unit 11 is completed, the output unit 32 displays the prediction results on the display unit (screen) of the terminal 40.
[0062] The format in which the prediction results are displayed is not particularly limited. For example, the prediction results may be displayed as a percentage (%) indicating the likelihood of developing a neoplastic disease. For example, it may be displayed as a numerical value such as "Likelihood of developing a neoplastic disease: 25%".
[0063] Furthermore, the prediction results may be displayed using visually intuitive symbols such as ○×△, where "×" indicates a high probability of infection, "△" indicates a moderate probability, and "○" indicates a low probability.
[0064] Furthermore, the prediction results may be displayed as levels. For example, they can be displayed in stages such as "Risk Level: High," "Risk Level: Medium," and "Risk Level: Low." Alternatively, a visually easy-to-understand display format may be adopted, such as a five-level rating (levels 1 to 5), a score display (out of 100 points, etc.), graphs, or meters.
[0065] Furthermore, the above display formats can be combined. For example, displaying both a percentage and a level (e.g., "Infection probability: 25% (risk level: low)") can improve the user's understanding of the information.
[0066] [Embodiment concerning proposal of preventive measures] In this embodiment, the disease prediction system preferably displays proposals for preventive measures for neoplastic diseases on the screen of the terminal 40 in conjunction with the output of the prediction results. The preventive measures are generated as content corresponding to the prediction results by the prediction unit 11 and are displayed on the screen of the user's terminal via the output unit 32.
[0067] The proposed preventive measures concern actions to prevent the onset of neoplastic diseases, and include, for example, dietary improvements (recommendations for foods containing antioxidants, intake of nutrients that boost immunity, etc.), administration of supplements (supplements with immune-boosting effects, vitamins, probiotics, etc.), administration of preventive medications, and lifestyle guidance to improve immunity (moderate exercise, stress reduction, regular health checkups, etc.).
[0068] Preferably, the proposed preventive measures display different content depending on the prediction result of the likelihood of developing a neoplastic disease by the prediction unit 11.
[0069] If the likelihood of developing a neoplastic disease is calculated to be low based on the fact that the person has valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes, it is preferable to display suggestions for preventive measures that are based on continuing the current lifestyle, while also including suggestions for preventive measures as a precaution, such as "Continue your current lifestyle," "It is recommended that you continue your current treatment and dietary management," "Please undergo regular health checkups as a precaution," and "Please try to eat a balanced diet to maintain your immunity."
[0070] Furthermore, even if a person has a history of valvular heart disease or similar conditions, if they are currently receiving treatment for that condition, additional information may be provided regarding the importance of continuing treatment and the prevention of complications related to the disease being treated.
[0071] On the other hand, if the likelihood of developing a neoplastic disease is calculated to be high based on the facts of vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis, it is preferable to display suggestions for preventive measures that include more proactive preventive measures. Specifically, for example, the following suggestions can be made.
[0072] (1) Suggestion to seek appropriate treatment for current vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis. Specifically, a message such as "Please seek appropriate treatment from a veterinarian for current vomiting, diarrhea, bloody stools (or allergic dermatitis, atopic dermatitis)" will be displayed. (2) Suggestion to have your pet examined by a veterinarian. Specifically, a message such as "We recommend seeing a veterinarian as soon as possible" or "Please have your pet undergo a comprehensive health checkup by a veterinarian for early detection of neoplastic diseases" will be displayed. (3) Suggestion regarding medication. Specifically, a message such as "Please consult with a veterinarian about medication to improve immunity" will be displayed.
[0073] Furthermore, if the likelihood of developing a neoplastic disease is calculated to be high, it may be indicated that treating current conditions such as vomiting, diarrhea, bloody stools, or allergic dermatitis may suppress chronic inflammation, resulting in improved immune function and a reduced risk of developing a neoplastic disease. This can motivate users (pet owners) by showing them that treating current conditions may also help prevent neoplastic diseases.
[0074] The system may be configured to extract and display appropriate suggestions from a database of preventive measures pre-stored in the memory unit 10, based on the prediction results and the input disease information. Alternatively, the prediction unit 11 may be configured to dynamically generate suggestions for preventive measures using a pre-trained model.
[0075] Please note that the suggested preventative measures are for general informational purposes only and are not intended as specific diagnoses or treatments. It is advisable to display a warning on the screen such as, "This suggestion is not a substitute for diagnosis or treatment by a veterinarian. For specific treatments, please consult your veterinarian."
[0076] [Other Embodiments of Disease Prediction Systems] Another embodiment of the disease prediction system of the present invention is characterized by comprising a correlation determination unit that compares the incidence rate of neoplastic diseases in animal groups that have suffered from a specific disease with the incidence rate of neoplastic diseases in animal groups that have not suffered from the specific disease, and determines that a given disease is correlated with neoplastic diseases if the incidence rates of neoplastic diseases differ between animal groups.
[0077] It is preferable that the incidence rate of neoplastic diseases in animal groups that have suffered from the aforementioned specific disease, and the incidence rate of neoplastic diseases in animal groups that have not suffered from the aforementioned specific disease, are calculated based on insurance claim data.
[0078] The correlation determination unit preferably uses a processor (arithmetic unit) such as a CPU or GPU in a computer or server to determine whether there is a correlation between the target disease and neoplastic diseases using a pre-set program, code, or software. The method of determination is not particularly limited, but as a preferred example, the processor uses a program that compares the incidence rate of neoplastic diseases in animal groups that have suffered from the specific disease with the incidence rate of neoplastic diseases in animal groups that have not suffered from the specific disease, performs a comparison, and preferably determines the significance of the difference, and determines that there is a correlation with neoplastic diseases for diseases where the incidence rate of neoplastic diseases differs between animal groups, preferably with a significant difference. Prevalence may be used instead of incidence rate. Incidence rate and prevalence rate can be calculated from a large database, such as a pet insurance database.
[0079] Other embodiments of the disease prediction system preferably include a prediction unit. The prediction unit is the same as that of the disease prediction system described above. In the case of other embodiments of the disease prediction system, the correlation determination unit, using the results of the correlation determination, receives information about the incidence of diseases correlated with neoplastic diseases, and uses that information to predict the incidence of neoplastic diseases, such as whether the risk of neoplastic disease is high or whether the probability of developing a neoplastic disease is low.
[0080] Furthermore, other embodiments of the disease prediction system may include a reception unit and an output unit.
[0081] [Insurance Premium Calculation System] The insurance premium calculation system of the present invention is characterized by comprising a prediction unit that uses information on the susceptibility of animals other than humans to specific diseases or benign tumors to predict whether or not an animal will contract a neoplastic disease, and an insurance premium calculation unit that calculates an insurance premium using the prediction results from the prediction unit. The prediction unit is the same as described above.
[0082] The premium calculation unit determines the insurance premium for the animal using the prediction results for the incidence of neoplastic disease from the prediction unit. Preferably, the premium calculation unit uses a processor (arithmetic unit) such as a CPU or GPU in a computer or server to calculate the insurance premium using the prediction results for the incidence of neoplastic disease, using a pre-set program, code, or software. The calculation method is not particularly limited, but a preferred example is a method in which the processor uses a program to adjust the insurance premium, which has been calculated according to a predetermined insurance premium table using basic information about the animal, such as age, sex, species, breed, weight, and information on the diversity of the gut microbiota, taking neoplastic disease into account. The insurance premium is increased if the risk of developing neoplastic disease is high, and decreased if the risk is low. In particular, regarding epilepsy, valvular heart disease, Malassezia otitis externa, and diabetes, a history of these diseases usually leads to higher insurance premiums. However, according to the present invention, a history of these diseases can be used to underestimate the risk of developing neoplastic diseases. Therefore, although other basic information also plays a role, it becomes possible to lower insurance premiums.
[0083] The determination of insurance premiums may involve setting a provisional premium based on the predicted incidence of neoplastic disease, and then revising that provisional premium using other information such as the animal's facial image, species, breed, age, sex, weight, medical history, and information on the diversity of its gut microbiota to calculate the final premium.
[0084] Furthermore, the present invention is not limited to a configuration comprising a prediction unit and an insurance premium calculation unit. For example, the insurance premium calculation unit may use information regarding the animal's susceptibility to a specific disease or benign tumor to calculate the insurance premium for the animal.
[0085] [Disease Prediction Method] The disease prediction method of the present invention is characterized by comprising a prediction step in which a computer uses information on the incidence of specific diseases or benign tumors in animals other than humans to predict whether or not an animal will contract a neoplastic disease. For the prediction step, refer to the prediction method by the prediction unit.
[0086] In the disease prediction method of the present invention, it is preferable that the information regarding the occurrence of a specific disease or benign tumor in an animal other than a human is information regarding whether or not the animal was affected within the past three years, and particularly within the past one year.
[0087] [Other Embodiments of Disease Prediction Methods] Another embodiment of the disease prediction method of the present invention is characterized by comprising a correlation determination step in which a computer compares the incidence rate of neoplastic diseases in an animal group that has suffered from a specific disease with the incidence rate of neoplastic diseases in an animal group that has not suffered from a specific disease, and determines that the predetermined disease is correlated with neoplastic diseases if the incidence rates of neoplastic diseases differ between the animal groups. The correlation determination step can refer to the determination made by the correlation determination unit described above.
[0088] In another aspect of the present invention, a disease prediction method is preferably one in which the incidence rate of neoplastic diseases in animal groups that have suffered from the specified disease, and the incidence rate of neoplastic diseases in animal groups that have not suffered from the specified disease, are calculated based on insurance claim data.
[0089] [Method for Calculating Insurance Premiums] The insurance premium calculation method of the present invention is characterized by comprising a prediction step in which a computer uses information on the susceptibility of animals other than humans to specific diseases or benign tumors to predict whether an animal will contract a neoplastic disease, and an insurance premium calculation step in which the computer calculates an insurance premium using the prediction result from the prediction step. The prediction step can refer to the prediction made by the prediction unit. The insurance premium calculation step can also refer to the calculation of insurance premiums by the insurance premium calculation unit.
[0090] Furthermore, the present invention is not limited to a configuration comprising a prediction step and an insurance premium calculation step. For example, in the insurance premium calculation step, information regarding the animal's susceptibility to a specific disease may be used to calculate the insurance premium.
[0091] [Death Prediction System] The death prediction system of the present invention is characterized by comprising a death prediction unit that uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will die within a predetermined period.
[0092] The mortality prediction unit of the present invention preferably uses a processor (arithmetic unit) such as a CPU or GPU in a computer or server to predict whether the target animal will die within a predetermined period of time, using a pre-set program, code, or software. The method of determination and prediction is not particularly limited, but a preferred example is that the processor uses information and data regarding the animal's susceptibility to benign tumors to predict whether it will die within a predetermined period of time.
[0093] In this invention, animals other than humans are preferably mammals, and dogs and cats are particularly preferred.
[0094] In this invention, the prediction of whether or not an animal will die refers to a prediction of whether or not the target animal is likely to die within a predetermined period, for example, within 5 years, 3 years, or 1 year from the time the prediction is made. The form of the prediction is not particularly limited and may include, for example, predicting whether or not an animal will die, indicating the probability of death, or determining the magnitude of the mortality risk. When indicating the probability of death, it is preferable to indicate it as a percentage between 0 and 100%. When determining the magnitude of the mortality risk, it is preferable to indicate the magnitude of the risk in stages, for example, large (high), medium, or small (low).
[0095] The mortality prediction unit preferably has a function to evaluate the probability of death based on whether or not an animal other than a human has ever suffered from a benign tumor. For example, if the animal has suffered from a benign tumor in the past, or has suffered from a benign tumor within a predetermined period preceding the prediction time, the probability of death is evaluated as high. Evaluating the probability of death as high means evaluating the probability of death as high or overestimating the risk of death. On the other hand, if the animal has never suffered from a benign tumor, or has not suffered from a benign tumor within a predetermined period preceding the prediction time, it is preferable to evaluate the probability of death as low. Evaluating the probability of death as low means evaluating the probability of death as low or underestimating the risk of death.
[0096] Furthermore, the mortality prediction unit can also use information regarding the occurrence of specific diseases to make predictions. For example, if there is a history of a specific disease in addition to a benign tumor, the risk of death may be estimated to be higher. The mortality prediction unit may also have a function to evaluate the likelihood of death depending on the type of specific disease. For example, the mortality prediction unit may evaluate the likelihood of death to be higher if an animal other than a human has a history of vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis. The prediction unit may also evaluate the likelihood of death to be lower if an animal has a history of valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes mellitus.
[0097] The prediction may use information other than information regarding the incidence of any specific disease in animals other than humans. Such information may include physical data such as the animal's age, breed, sex, weight, height, and length, as well as sequence data such as genome sequences, SNPs, and specific gene sequences, medical history, family history, and information regarding the diversity of the gut microbiota. Preferably, information regarding weight is used for prediction. The inventors have found a correlation between the weight of an animal and the incidence of neoplastic diseases. By utilizing this correlation, it is expected that the accuracy of predicting whether an animal will die can be improved by using information regarding the weight of the animal being predicted. It is preferable to predict that the higher the weight of the animal being predicted, the greater the risk of death. As information regarding weight, the numerical value of the animal's weight may be used, or a classification that divides the weight into predetermined ranges may be used. A classification that divides the weight into predetermined ranges may, for example, be divided into large, medium, small, and very small according to weight, and information on which classification the animal belongs to is used for prediction. It is also preferable to use information regarding the diversity of the gut microbiota. Information on the diversity of the gut microbiota includes the Shannon Index (Shannon-Wiener diversity index) and the Simpson Diversity Index. Information on the diversity of the gut microbiota can be obtained using known metagenomic analysis methods such as amplicon sequencing using sequencers such as NGS, or microbiota analysis methods, using samples such as feces collected from the target animals.
[0098] Information regarding the incidence of benign tumors in animals other than humans is the same as above.
[0099] Furthermore, the mortality prediction unit may perform predictions and decisions using a pre-trained model. Examples of pre-trained models include a model that has learned the relationship between information on the occurrence of benign tumors and whether or not death occurred within a predetermined period. Such a pre-trained model can be obtained, for example, by training it with information on whether an animal was diagnosed with a benign tumor within a predetermined period and whether or not that animal subsequently died as training data. Whether or not death occurred can be replaced with a dummy variable. Information on whether or not the animal died can be obtained, for example, from a veterinary hospital or an insured pet owner as an insurance claim (also called an "accident").
[0100] The mortality prediction system of the present invention may also include a prediction unit, a reception unit, an output unit, and other components. Details of the prediction unit, reception unit, and output unit are the same as described above.
[0101] [Further Embodiments of Disease Prediction Systems] The present invention includes, as yet another embodiment of a disease prediction system, a disease prediction system comprising a kidney disease prediction unit that predicts whether an animal other than a human will develop chronic kidney disease, using information on the incidence of benign tumors in such animals.
[0102] The kidney disease prediction unit of the present invention preferably uses a processor (arithmetic unit) such as a CPU or GPU in a computer or server to predict whether a target animal will develop chronic kidney disease using a pre-set program, code, or software. The method of determination and prediction is not particularly limited, but a preferred example is that the processor uses information and data regarding the target animal's susceptibility to benign tumors to predict whether it will develop chronic kidney disease.
[0103] In the disease prediction system equipped with the kidney disease prediction unit of the present invention, mammals are preferred as animals other than humans, dogs and cats are more preferred, and cats are particularly preferred. Cats have a high mortality rate from chronic kidney disease, and chronic kidney disease, along with malignant tumors, is a serious disease for cats. The present invention makes it possible to predict the risk of kidney disease in cats and other animals, and based on this prediction, it becomes possible to propose measures to reduce the risk of chronic kidney disease, such as changing to a diet with low protein, phosphorus, or sodium content that is easy on the kidneys, or increasing water intake.
[0104] In the disease prediction system equipped with the kidney prediction unit of the present invention, the prediction refers to the prediction of whether the target animal will develop chronic kidney disease within a predetermined period, for example, within 5 years, 3 years, or 1 year from the time the prediction is performed. The form of the prediction is not particularly limited, and examples include predicting whether or not an animal will develop chronic kidney disease, indicating the probability of developing the disease, or determining the magnitude of the risk. When indicating the probability of developing chronic kidney disease, it is preferable to indicate it as a percentage between 0 and 100%. When determining the magnitude of the risk, it is preferable to indicate the magnitude of the risk in stages, for example, large (high), medium, or small (low).
[0105] The kidney disease prediction unit preferably has a function to evaluate the likelihood of developing chronic kidney disease based on whether or not an animal other than a human has ever had a benign tumor. For example, if the animal has had a benign tumor in the past, or has had a benign tumor within a predetermined period preceding the prediction time, the likelihood of developing chronic kidney disease should be evaluated as high. Evaluating the likelihood of developing chronic kidney disease as high means evaluating the probability of developing the disease as high or overestimating the risk of developing the disease. On the other hand, if the animal has never had a benign tumor, or has not had a benign tumor within a predetermined period preceding the prediction time, it is preferable to evaluate the likelihood of developing chronic kidney disease as low. Evaluating the likelihood of developing chronic kidney disease as low means evaluating the probability of developing the disease as low or underestimating the risk of developing the disease.
[0106] Furthermore, the kidney disease prediction unit can also make predictions using information regarding the occurrence of specific diseases. For example, if there is a history of a specific disease in addition to a benign tumor, the risk of developing chronic kidney disease may be estimated to be higher. The kidney disease prediction unit may also have a function to evaluate the likelihood of developing chronic kidney disease depending on the type of specific disease. For example, the kidney disease prediction unit may evaluate the likelihood of developing chronic kidney disease to be higher if an animal other than a human has a history of vomiting, diarrhea, bloody stools, allergies, or atopic dermatitis. The kidney disease prediction unit may also evaluate the likelihood of developing chronic kidney disease to be lower if an animal has a history of valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes mellitus.
[0107] To predict the incidence of chronic kidney disease, information other than that related to the incidence of any specific disease in animals other than humans may be used. Such information may include physical data such as the animal's age, breed, sex, weight, height, and length; sequence data such as genome sequences, SNPs, and specific gene sequences; medical history; family history; and information on the diversity of the gut microbiota. Preferably, information related to weight is used for prediction. It is preferable to predict that the higher the weight of the animal being predicted, the greater the risk of developing chronic kidney disease. As information related to weight, the numerical value of the animal's weight may be used, or a classification that divides weight into predetermined ranges may be used. A classification that divides weight into predetermined ranges may, for example, be divided into large, medium, small, and very small according to weight, and information on which classification the animal belongs to is used for prediction. It is also preferable to use information related to the diversity of the gut microbiota. Examples of information related to the diversity of the gut microbiota include the Shannon Index (Shannon-Wiener diversity index) and the Simpson Diversity Index. Information on the diversity of the gut microbiota can be obtained using known metagenomic analysis methods such as amplicon sequencing using sequencers such as NGS, or bacterial analysis methods, with samples such as feces collected from the animals to be predicted.
[0108] Information regarding the incidence of benign tumors in animals other than humans refers to information on whether or not an animal has had a benign tumor in the past, preferably information on whether or not the animal had a benign tumor before the prediction unit performed the prediction. More preferably, it refers to information on whether or not the animal had a benign tumor within 10 years, 5 years, 3 years, or 1 year prior to the prediction unit performed the prediction. The form and input method of the information are not particularly limited, and known forms and input methods can be adopted.
[0109] The method for obtaining information on benign tumors in animals other than humans, or on the occurrence of any of the specified diseases, is not particularly limited. For example, it can be obtained and entered through pet insurance claim records, input by users of the system of the present invention, declarations, uploads, etc. For example, one method is to upload information on the occurrence of benign tumors, along with basic information such as the age, sex, species, and breed of the animal to be predicted, from a terminal used by the user to the system of the present invention via an internet connection.
[0110] Furthermore, the kidney disease prediction unit may perform predictions and judgments using a pre-trained model. Examples of pre-trained models include a model that has learned the relationship between information on the occurrence of benign tumors and the presence or absence of chronic kidney disease. Such a pre-trained model can be obtained, for example, by training it with information on whether an animal was diagnosed with a benign tumor within a certain period, and whether that animal subsequently developed chronic kidney disease, as training data. Whether or not an animal was diagnosed with a disease can be replaced with a dummy variable. Information on whether or not an animal was diagnosed with a disease can be obtained, for example, from a veterinary hospital or the insured owner as an insurance claim (also called an "accident"). In other words, if the animal is insured as a pet, and the animal is diagnosed with a disease after visiting a hospital, the veterinary hospital or the owner (pet insurance policyholder) will claim insurance payment from the insurance company along with the fact of the disease, so the insurance company can learn that the animal was diagnosed with a disease. On the other hand, if no claim for insurance benefits is made within a specified period of time, it can be determined that the insured animal did not suffer from the disease during that period.
[0111] A disease prediction system according to yet another aspect of the present invention may also include a prediction unit, a reception unit, an output unit, and other components. Details of the prediction unit, reception unit, and output unit are the same as described above.
[0112] [Reference Example 1] By referring to the pet insurance claim database, we investigated whether any of the dogs for which an insurance claim was made due to a neoplastic disease had not had an insurance claim made for a neoplastic disease in the previous year, and whether they had a history of other diseases in the year prior to the insurance claim in question.
[0113] A comparison of groups with a history of other diseases (the group with a history of disease) and groups without (the group without incidents and the group with a history of insurance claims due to other diseases) revealed that for many diseases, the group with a history of disease tended to have a higher incidence (prevalence) of neoplastic diseases. This is likely because animals suffering from any disease are in a state of reduced health, and their immune function, such as the immune mechanism that suppresses neoplastic diseases, may have been impaired.
[0114] As shown in Figures 3 to 5, the prevalence of neoplastic diseases tended to be higher in groups with a history of vomiting, diarrhea, and bloody stools (Figure 3), atopic dermatitis (Figure 4), and allergic dermatitis (Figure 5) compared to groups without such a history. In addition, although not shown in the figures, the prevalence of neoplastic diseases tended to be higher in groups with a history of periodontal disease, biliary sludge, gastritis and enteritis, as well as symptoms such as lethargy.
[0115] Furthermore, as shown in Figure 6, a higher incidence (prevalence) of neoplastic diseases was observed in the group with a history of skin tumors (lipomas and histiocytomas). Generally, these skin tumors are considered benign tumors and are distinguished from malignant tumors, and the relationship between benign and malignant tumors has not been deeply explored until now. However, these results suggest that aging and immune function are involved in benign tumors as well, and that not all benign tumors arise from selective tolerance by the immune system, but rather that pathological development may be caused by immunosuppression, such as immaturity or exhaustion of the immune system. In other words, it is thought that benign and malignant tumors may share common mechanisms at the cellular development stage, and that immune overreaction (immaturity, exhaustion, and carcinogenesis) may be a branching point between the two and contribute to the prediction of neoplastic diseases.
[0116] Therefore, even if an animal does not currently have a neoplastic disease, it may be considered highly likely (high risk) of developing one in the future due to other underlying health conditions. As a result, when calculating the insurance premium for such an animal, it may be appropriate to set a higher premium.
[0117] In contrast, diseases such as valvular heart disease showed a different trend from that of vomiting, diarrhea, bloody stools, and atopic dermatitis. As shown in Figures 7 and 8, the group with a history of valvular heart disease (Figure 7) and epilepsy (Figure 8) tended to have a lower incidence (prevalence) of neoplastic diseases compared to the group without a history of these conditions (insurance claims made for other diseases). Although not shown in the figures, the group with a history of Malassezia otitis externa and diabetes also tended to have a lower incidence (prevalence) of neoplastic diseases compared to the group without a history of these conditions (insurance claims made for other diseases). One possible reason for this is that, for certain diseases like these, the animals' health improved as a result of treatment and dietary care for the disease, such as an increase in the diversity of the gut microbiota, which may have made them less susceptible to neoplastic diseases. In other words, it is possible that the immune function, such as the function of the immune mechanism that suppresses neoplastic diseases, improved.
[0118] Therefore, for certain diseases, it is conceivable to assess that the likelihood of developing a neoplastic disease in the future is low (the risk is low) based on the fact that the animal currently has (or has a history of having) that disease. As a result, when calculating the insurance premium for the animal, the premium may be calculated at a lower rate.
[0119] For diseases other than those mentioned above, by comparing the incidence or prevalence of neoplastic diseases with and without the presence of the disease in question, if a difference in incidence can be found, the likelihood of developing a neoplastic disease can be assessed accordingly. Furthermore, the insurance premium for the animal in question can be calculated accordingly (higher or lower).
[0120] [Reference Example 2] Referring to the pet insurance claim database, individuals (dogs) for which insurance claims were made due to neoplastic diseases were divided into a mixed breed group and a purebred dog group. For each group, individuals that had not made insurance claims due to neoplastic diseases in the previous year were investigated to see whether there had been an insurance claim for a benign tumor (lipoma or histiocytoma among skin tumors; unless otherwise specified, in the examples or reference examples, "benign tumor" refers to lipoma or histiocytoma among skin tumors) in the year prior to the insurance claim. Mixed breed dogs are dogs whose breed registration at the time of insurance contract is "mixed breed," and purebred dogs are dogs other than those whose breed registration is "mixed breed." The results are shown in Figure 9.
[0121] As can be seen in Figure 9, in both the mixed-breed and purebred groups, when comparing the group with a history of benign tumors (group with two types of benign tumors in the previous year) and the group without such a history (no incidents in the previous year and no benign tumors in the previous year), a tendency was observed for the group with a history of benign tumors to have a higher incidence (prevalence) of neoplastic diseases.
[0122] [Reference Example 3] By referring to the pet insurance claim database, we collected individuals for whom insurance claims were made due to neoplastic diseases excluding skin tumors. This group was then divided into three groups: those for whom insurance claims were made for benign tumors in the previous year, those for whom no claims were made, and those for whom there were no claims in the previous year. The mortality rate for each group was then calculated. "Mortality rate" refers to the percentage of pet insurance policies that were canceled due to the death of the insured pet. The death cancellation rate is calculated as "number of death cancellations" / "number of policies" during a specified period, and this is defined as the "mortality rate." The results are shown in Figure 10.
[0123] As can be seen from Figure 10, the group with a history of benign tumors (the group with two types of benign tumors in the previous year) had a higher mortality rate than the group with no history of benign tumors or the group with no prior incidents. This indicates that the group that developed a malignant tumor after having a benign tumor had a higher mortality rate than the group that developed a malignant tumor without having a history of a benign tumor.
[0124] [Reference Example 4] Referring to the pet insurance claim database, individuals (dogs) for which insurance claims were made due to benign tumors were divided into a mixed breed group and a purebred dog group, and the age and incidence of benign tumors were investigated for each group. The results are shown in Figure 11.
[0125] Figure 11 shows that the incidence of benign tumors increases with age in both mixed-race and purebred groups.
[0126] [Reference Example 5] Referring to the pet insurance claims database, individual dogs were divided into large dogs (20 kg or more), medium-sized dogs (10 kg or more but less than 20 kg), and small dogs (less than 10 kg) according to their weight, and the incidence of benign tumors by age was investigated for each group. The results are shown in Figure 12.
[0127] Figure 12 shows that large dogs have a higher incidence of benign tumors than other groups.
[0128] [Reference Example 6] Referring to the pet insurance claims database, mixed breed dogs were divided into large breeds (over 20 kg), medium breeds (10 kg to under 20 kg), and small breeds (under 10 kg) according to their weight, and the incidence of benign tumors by age was investigated for each group. The results are shown in Figure 13.
[0129] Figure 13 shows that large dogs have a higher incidence of benign tumors than other groups.
[0130] [Reference Example 7] Referring to a pet insurance claims database, poodles were divided into poodles (standard) and poodles (toy poodles and miniature poodles) according to their body size, and the incidence of benign tumors by age was investigated for each group. The results are shown in Figure 14.
[0131] Figure 14 shows that even among poodles, standard poodles, which are larger in size, have a higher incidence of benign tumors than toy poodles and miniature poodles.
[0132] [Reference Example 8] By referring to the pet insurance claim database, dogs were divided into two groups: those with poor gut health and those with good gut health. The division into good / bad gut health was as follows: Pet insurance policyholders sent fecal samples of their insured dogs to the insurance company, and the Shannon Index of the dog's gut microbiota was calculated based on the results of metagenomic analysis using a next-generation sequencer. Then, each dog was divided into large dogs (over 20 kg), medium dogs (10 kg to under 20 kg), and small dogs (under 10 kg), and the standard score of the Shannon Index was calculated for each group. For small dogs, those with a diversity standard score of less than 42 were classified as having "poor gut health," and those with a diversity standard score of 42 or higher were classified as having "good gut health." For medium-sized dogs, those with a diversity standard score of less than 46 were classified as having "poor gut health," and those with a diversity standard score of 46 or higher were classified as having "good gut health." For large breed dogs, those with a diversity deviation score of less than 50 were classified as having a "poor gut environment," and those with a diversity deviation score of 50 or higher were classified as having a "good gut environment." The incidence of benign tumors within one year of fecal sample collection was compared between the "poor gut environment" group and the "good gut environment" group of large, medium, and small breed dogs. The results are shown in Figure 15. The analysis included only dogs that had been accident-free for 90 days prior to the analysis of fecal samples for gut microbiota analysis.
[0133] Figure 15 shows that the group with poor gut health had a higher incidence of benign tumors than the group with good gut health.
[0134] [Reference Example 9] Referring to a pet insurance claims database, dogs were divided into two groups: one with poor gut health and one with good gut health, in the same manner as described above. The incidence of lipomas (a type of benign tumor) within one year of fecal sample collection was then compared between the "poor gut health" group and the "good gut health" group. The results are shown in Figure 16. Note that the analysis was limited to dogs that had been accident-free for 90 days prior to the analysis of fecal samples for gut microbiota analysis.
[0135] Figure 16 shows that the group with poor gut health had a higher incidence of lipomas than the group with good gut health.
[0136] [Reference Example 10] Referring to a pet insurance claims database, dogs were divided into two groups: one with poor gut health and one with good gut health, in the same manner as described above. The incidence of histiocytoma (a type of benign tumor) within one year of fecal sample collection was then compared between the "poor gut health" group and the "good gut health" group. The results are shown in Figure 17. Note that the analysis was limited to dogs that had been accident-free for 90 days prior to the analysis of fecal samples for gut microbiota analysis.
[0137] Figure 17 shows that the group with a poor gut environment had a higher incidence of histiocytoma than the group with a good gut environment.
[0138] [Reference Example 11] Referring to the pet insurance claims database, mixed breed dogs were divided into large breeds (20 kg or more), medium breeds (10 kg to less than 20 kg), and small breeds (less than 10 kg) based on weight, and the incidence of malignant neoplastic diseases by age in each group was examined. The results are shown in Figure 18.
[0139] Figure 18 shows that the group with higher body weight had a higher incidence of neoplastic diseases than the group with lower body weight.
[0140] [Reference Example 12] Referring to the pet insurance claims database, the Shannon Index of the canine gut microbiota was calculated in the same manner as above. The diversity deviation score was calculated for each group of large, medium, and small dogs, and the dogs were divided into "low diversity," "moderate diversity," and "high diversity" groups according to their diversity deviation scores. For each group, the incidence of neoplastic diseases within one year from the time of fecal sample collection was examined. The results are shown in Figure 19. Note that the analysis was limited to dogs that had been accident-free for 90 days prior to the analysis of the fecal sample for gut microbiota analysis.
[0141] Figure 19 shows that the group with low diversity in the gut microbiota had a higher incidence of neoplastic diseases than the group with moderate or high diversity.
[0142] [Reference Example 13] Referring to a pet insurance claims database, dogs were divided into a group with poor gut health and a group with good gut health in the same manner as described above. The incidence of lymphoma (cutaneous lymphoma, gastrointestinal lymphoma, multicentric lymphoma) within one year of fecal sample collection was then compared between the "poor gut health" group and the "good gut health" group. The results are shown in Figure 20. Note that individuals with a history of lymphoma were excluded from the analysis of fecal samples for gut microbiota analysis.
[0143] Figure 20 shows that the group with poor gut health had a higher incidence of lymphoma than the group with good gut health.
[0144] [Reference Example 14] Figure 21 shows an example of how to score the risk of developing neoplastic diseases in dogs, based on whether or not an insurance claim was made, whether or not an insurance claim was made due to a benign tumor, and the results of the gut microbiota analysis (high or low diversity index) in the following year. According to Figure 21, individuals that did not have an insurance claim (incident) in the first year and had a high gut microbiota diversity index in the following year are assigned a score of "0" as the risk of developing neoplastic diseases. Individuals that did not have an insurance claim (incident) in the first year and had a low gut microbiota diversity index in the following year are assigned a score of "2.5" as the risk of developing neoplastic diseases. In addition, individuals that had an insurance claim (incident other than a benign tumor) in the first year and had a high gut microbiota diversity index in the following year are assigned a score of "5" as the risk of developing neoplastic diseases. Individuals who file an insurance claim (for an incident other than a benign tumor) in the first year and have a low gut microbiota diversity index in the following year are assigned a score of "7.5" as their risk of developing a neoplastic disease. Individuals who file an insurance claim for a benign tumor in the first year and have a high gut microbiota diversity index in the following year are assigned a score of "15" as their risk of developing a neoplastic disease. Individuals who file an insurance claim for a benign tumor in the first year and have a low gut microbiota diversity index in the following year are assigned a score of "17.5" as their risk of developing a neoplastic disease. Thus, in the example in Figure 21, a low diversity index in the current year increases the risk assessment score by 2.5, an insurance claim for a disease other than a benign tumor in the previous year increases the risk assessment score by 5, and an insurance claim for a benign tumor in the previous year increases the risk assessment score by 15. The magnitude of this increase in the risk score is not particularly limited and may be set as appropriate. Preferably, the increase in the risk assessment score is increased in the following order: the fact of an insurance claim due to a benign tumor in the previous year, the fact of an insurance claim due to a disease other than a benign tumor in the previous year, and a low diversity index for the current year (the fact of an insurance claim due to a benign tumor in the previous year results in the largest increase).
[0145] Figure 22 is a graph comparing the incidence of neoplastic diseases in dogs, based on a pet insurance claims database, where each dog was divided into groups corresponding to the grouping in Figure 21. This shows that the group with the highest risk assessment score in Figure 21—those with a benign tumor claim in the previous year and a low diversity index—has the highest incidence of neoplastic diseases.
[0146] Figure 23 is a graph showing the correlation between the risk assessment score values in Figure 21 and the incidence rate of neoplastic diseases. From Figure 23, a correlation can be seen between the risk assessment score values set in Figure 21 and the incidence rate of neoplastic diseases, indicating that the risk assessment score values in Figure 21 were appropriate.
[0147] [Reference Example 15] Figure 24 shows an example of how to score the risk of developing chronic kidney disease in cats, based on whether or not an insurance claim was filed, whether or not an insurance claim was filed due to a benign tumor, and the results of the gut microbiota analysis (high or low diversity index) in the following year. According to Figure 24, individuals that did not file an insurance claim (incident) in the first year and had a high gut microbiota diversity index in the following year are assigned a score of "0" as the risk of developing chronic kidney disease. Individuals that did not file an insurance claim (incident) in the first year and had a low gut microbiota diversity index in the following year are assigned a score of "5" as the risk of developing chronic kidney disease. In addition, individuals that filed an insurance claim (incident other than a benign tumor) in the first year and had a high gut microbiota diversity index in the following year are assigned a score of "2.5" as the risk of developing chronic kidney disease. Individuals who file an insurance claim (for an incident other than a benign tumor) in the first year and have a low gut microbiota diversity index in the following year are assigned a score of "7.5" as a risk factor for developing chronic kidney disease. Individuals who file an insurance claim for a benign tumor in the first year are assigned a score of "20". Thus, in the example in Figure 24, a low diversity index in the current year increases the risk assessment score by 5, an insurance claim for a disease other than a benign tumor in the previous year increases the risk assessment score by 2.5, and an insurance claim for a benign tumor in the previous year increases the risk assessment score by 20. The magnitude of this risk increase score is not particularly limited and may be set as appropriate. Preferably, the increase in the risk assessment score should be increased in the following order: the fact of an insurance claim for a benign tumor in the previous year, the fact of an insurance claim for a disease other than a benign tumor in the previous year, and a low diversity index in the current year (the fact of an insurance claim for a benign tumor in the previous year results in the largest increase).
[0148] Figure 25 is a graph comparing the prevalence of chronic kidney disease in cats, divided into groups corresponding to the groupings in Figure 24, based on a pet insurance claims database. This shows that the group with the highest risk assessment score in Figure 24—those with a benign tumor claim in the previous year—has the highest prevalence of chronic kidney disease.
[0149] Figure 26 is a graph showing the correlation between the risk assessment score values in Figure 24 and the prevalence of chronic kidney disease. From Figure 26, a correlation can be seen between the risk assessment score values set in Figure 24 and the prevalence of chronic kidney disease, indicating that the risk assessment score values in Figure 24 were appropriate.
[0150] [Reference Example 16] Referring to the pet insurance claim database, we examined the prevalence of neoplastic diseases (insurance claim rates due to neoplastic diseases) for cats that had made insurance claims due to neoplastic diseases. The data was divided into three groups: a group with no claims in the previous year (no claims in the previous year), a group with claims due to six types of benign skin tumors formed in the previous year (specifically, skin tumors with undetermined pathology, lipoma, histiocytoma (skin), mast cell tumor (skin), perianal tumor, and other benign skin tumors) (six types of benign tumors in the previous year), and a group with claims for reasons other than the six types of benign tumors in the previous year (other claims in the previous year). The results are shown in Figure 27.
[0151] As can be seen from Figure 27, in each age group, the incidence of neoplastic diseases was highest in the group that had a benign tumor (6 types) in the previous year, followed by the group that had an accident other than a benign tumor (6 types) in the previous year.
[0152] [Reference Example 17] Referring to a pet insurance claims database, cats were divided into large (6.0 kg or more), medium, and small (less than 6.0 kg) groups according to their weight, and the incidence of neoplastic diseases by age was investigated for each group. The results are shown in Figure 28.
[0153] Figure 28 shows that large cats have a higher incidence of neoplastic diseases than medium-sized and small cats.
[0154] [Reference Example 18] Referring to a pet insurance claims database, cats were divided into large (6.0 kg or more), medium, and small (less than 6.0 kg) groups according to their weight, and the incidence of two types of benign tumors (lipoma and histiocytoma) by age was investigated for each group. The results are shown in Figure 29.
[0155] Figure 29 shows that large cats have a higher incidence of two types of benign tumors (lipoma and histiocytoma) than medium-sized and small cats.
[0156] [Reference Example 19] Referring to a pet insurance claims database, cats were divided into large (6.0 kg or more), medium, and small (less than 6.0 kg) groups according to their weight, and the incidence of skin tumors (including benign skin tumors and malignant neoplastic diseases such as melanoma, cytoplasmosis, and cutaneous lymphoma) by age was investigated for each group. The results are shown in Figure 30.
[0157] Figure 30 shows that the incidence of skin tumors is higher in large cats than in medium and small cats.
[0158] [Reference Example 20] Referring to a pet insurance claims database, cats were divided into large (6.0 kg or more), medium, and small (less than 6.0 kg) groups according to their weight, and the incidence of neoplastic diseases other than skin tumors (neoplastic diseases excluding melanoma, melanocytoma, and cutaneous lymphoma) was investigated for each group by age. The results are shown in Figure 31.
[0159] Figure 31 shows that large cats have a higher incidence of neoplastic diseases other than skin tumors than medium-sized and small cats.
[0160] The present invention includes the following [1] to
[16] : [1] A disease prediction system characterized by comprising a prediction unit that predicts whether an animal other than a human will contract a neoplastic disease, using information on the animal's susceptibility to valvular heart disease, epilepsy, vomiting / diarrhea / bloody stool disease, atopic dermatitis, or allergic dermatitis. [2] The disease prediction system of [1], further comprising a reception unit that receives information on the animal other than a human's susceptibility to valvular heart disease, epilepsy, vomiting / diarrhea / bloody stool disease, atopic dermatitis, or allergic dermatitis. [3] The disease prediction system of [1], wherein the prediction unit evaluates the likelihood of the animal other than a human contracting a neoplastic disease to be high if the animal has previously suffered from vomiting / diarrhea / bloody stool disease, allergic dermatitis, or atopic dermatitis. [4] The disease prediction system of [1], wherein the prediction unit evaluates the likelihood of the animal other than a human contracting a neoplastic disease to be low if the animal has previously suffered from valvular heart disease or epilepsy. [5] The disease prediction system of [1] in which the information relating to the incidence of valvular heart disease, epilepsy, vomiting / diarrhea / bloody stool disease, atopic dermatitis, or allergic dermatitis in animals other than humans is information relating to whether or not the animals have suffered from valvular heart disease, epilepsy, vomiting / diarrhea / bloody stool disease, atopic dermatitis, or allergic dermatitis within the past three years. [6] A disease prediction system characterized by comprising a correlation determination unit that compares the incidence of neoplastic diseases in an animal group that has suffered from a specified disease excluding neoplastic diseases with the incidence of neoplastic diseases in an animal group that has never suffered from a specified disease excluding neoplastic diseases, and determines that the specified disease is correlated with neoplastic diseases if the incidence of neoplastic diseases differs between the animal groups. [7] The disease prediction system of [6] in which the incidence of neoplastic diseases in an animal group that has suffered from a specified disease excluding neoplastic diseases, and the incidence of neoplastic diseases in an animal group that has never suffered from a specified disease excluding neoplastic diseases, are calculated based on insurance claim data. [8] An insurance premium calculation system comprising: a prediction unit that predicts whether an animal will contract a neoplastic disease using information on the incidence of valvular heart disease, epilepsy, vomiting, diarrhea, bloody stool disease, atopic dermatitis, or allergic dermatitis in animals other than humans; and an insurance premium calculation unit that calculates an insurance premium using the prediction results from the prediction unit.[9] A disease prediction method characterized by comprising a prediction step in which a computer predicts whether an animal other than a human will contract a neoplastic disease, using information on the animal's susceptibility to valvular heart disease, epilepsy, vomiting, diarrhea, bloody stools, atopic dermatitis, or allergic dermatitis.
[10] The disease prediction method according to [9], further comprising a step of receiving information on the animal other than a human's susceptibility to valvular heart disease, epilepsy, vomiting, diarrhea, bloody stools, atopic dermatitis, or allergic dermatitis.
[11] The disease prediction method according to [9], wherein the prediction step evaluates the likelihood of the animal other than a human contracting a neoplastic disease as high if the animal has previously contracted vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis.
[12] The disease prediction method according to [9], wherein the prediction step evaluates the likelihood of the animal contracting a neoplastic disease as low if the animal other than a human has previously contracted valvular heart disease or epilepsy.
[13] A disease prediction method for animals other than humans, wherein the information relating to the incidence of valvular heart disease, epilepsy, vomiting, diarrhea, bloody stool disease, atopic dermatitis, or allergic dermatitis is information relating to whether the animals have suffered from valvular heart disease, epilepsy, vomiting, diarrhea, bloody stool disease, atopic dermatitis, or allergic dermatitis within the past three years.
[14] A disease prediction method characterized by comprising a correlation determination step in which a computer compares the incidence rate of neoplastic diseases in an animal group that has suffered from a predetermined disease excluding neoplastic diseases with the incidence rate of neoplastic diseases in an animal group that has never suffered from a predetermined disease excluding neoplastic diseases, and determines that the predetermined disease is correlated with neoplastic diseases if the incidence rates of neoplastic diseases differ between the animal groups.
[15] A disease prediction method according to
[14] , wherein the incidence rate of neoplastic diseases in an animal group that has suffered from a predetermined disease other than neoplastic diseases, and the incidence rate of neoplastic diseases in an animal group that has never suffered from a predetermined disease other than neoplastic diseases, are calculated based on insurance claim data.
[16] An insurance premium calculation method characterized by comprising a prediction step in which a computer uses information on the incidence of valvular heart disease, epilepsy, vomiting, diarrhea, bloody stool disease, atopic dermatitis, or allergic dermatitis in animals other than humans to predict whether the animal will suffer from a neoplastic disease, and an insurance premium calculation step in which the insurance premium is calculated using the prediction result from the prediction step.
Claims
1. A disease prediction system characterized by comprising a prediction unit that predicts whether an animal will contract a neoplastic disease, using information on the incidence of a specified disease, excluding neoplastic diseases, in animals other than humans.
2. The disease prediction system according to claim 1, wherein the prediction unit evaluates the likelihood of the animal (excluding humans) developing a neoplastic disease to be low if the animal has previously suffered from at least one of the following: valvular heart disease, epilepsy, Malassezia otitis externa, or diabetes mellitus.
3. A disease prediction system characterized by comprising a correlation determination unit that compares the incidence rate of neoplastic diseases in an animal group that has suffered from a specified disease excluding neoplastic diseases with the incidence rate of neoplastic diseases in an animal group that has never suffered from a specified disease excluding neoplastic diseases, and determines that the specified disease is correlated with neoplastic diseases if the incidence rates of neoplastic diseases differ between the animal groups.
4. An insurance premium calculation system characterized by comprising an insurance premium calculation unit that calculates an insurance premium for an animal other than a human, using information regarding the animal's susceptibility to a specified disease, excluding neoplastic diseases.
5. A disease prediction method characterized by comprising a prediction step in which a computer uses information on the susceptibility of animals other than humans to a predetermined disease, excluding neoplastic diseases, to predict whether the animal will contract a neoplastic disease.
6. A method for calculating insurance premiums, characterized in that a computer calculates an insurance premium for an animal other than a human, using information regarding the animal's susceptibility to a predetermined disease, excluding neoplastic diseases.
7. The disease prediction system according to claim 1, wherein the prediction unit evaluates the likelihood of a non-human animal developing a neoplastic disease if the animal has previously suffered from at least one of the following: vomiting, diarrhea, bloody stools, allergic dermatitis, or atopic dermatitis.
8. A disease prediction system characterized by comprising a prediction unit that predicts whether an animal will develop a neoplastic disease, using information on the incidence of benign tumors in animals other than humans.
9. The disease prediction system according to claim 8, wherein the prediction unit evaluates the likelihood of an animal other than a human being developing a neoplastic disease if the animal has previously suffered from a benign tumor.
10. The disease prediction system according to claim 8, wherein the prediction unit further uses information regarding the animal's weight to predict whether the animal will suffer from a neoplastic disease.
11. The disease prediction system according to claim 8, wherein the prediction unit further predicts whether the animal will suffer from a neoplastic disease using information regarding the diversity of the animal's gut microbiota.
12. A mortality prediction system characterized by comprising a mortality prediction unit that predicts whether an animal other than a human will die within a predetermined period of time, using information on the incidence of benign tumors in the animal.
13. A disease prediction system characterized by comprising a kidney disease prediction unit that predicts whether an animal will develop chronic kidney disease, using information on the incidence of benign tumors in animals other than humans.
14. An insurance premium calculation system characterized by comprising an insurance premium calculation unit that calculates the insurance premium for an animal other than a human using information on the incidence of benign tumors in the animal.
15. A disease prediction method characterized by comprising a prediction step in which a computer uses information on the incidence of benign tumors in animals other than humans to predict whether the animal will develop a neoplastic disease.
16. A method for calculating insurance premiums, characterized in that a computer includes a premium calculation step for calculating the insurance premium of an animal other than a human, using information regarding the incidence of benign tumors in the animal.
