2017 Volume 43 Issue 2 Pages 117-122
Foot-and-mouth disease (FMD), a “specific domestic animal infectious disease,” needs biosecurity in the early stage but cannot be detected during the incubation period. This problem is dealt with by use of a detection agent and by immediate culling when the outbreak begins. However, developing a detection agent raises budgetary problems, and securing sites for culling and burial takes time. In this study, we adopted an epidemiological model known as the Keeling Model and developed a mathematical method to detect infected farms during the incubation period before clinical symptoms first appear. Our model based on the Keeling Model estimates the unobservable variate λ (the infection index), which indicates the situation of the susceptible farms with the observable variates of whether or not surrounding farms infected, the number and species of animals, and the distances between farms. In addition, we devised a model (Infection Estimating Model) and an estimation approach to estimate whether or not a farm is infected. The proposed method was applied to the case of FMD that occurred in Miyazaki Prefecture in 2010, and the validity of our proposal was confirmed. Moreover, as a result of an experiment to compare estimation accuracy, a universal factor was discovered that will improve estimation accuracy.