DEVELOPMENT OF A PRECISION EPIDEMIOLOGY WEB-BASED TOOL FOR LIVESTOCK DISEASE MANAGEMENT

<b>Project Methods</b><br> Objective 1: Develop and implement a web-service tool enabling the automatic connection,integration, secure access, and multi-level visualization of key livestock-health data.The approach is to develop standard routines, web-services anda user-friendly interface that allow producers and veterinary practitioners to upload, integrate,search, query and visualize the available data (i.e., diagnostic data, feed management, productionrecords, biosecurity and management practices, environmental and climatic information)regarding their animal populations and their premises at any time.Objective 2: Develop an innovative precision epidemiology (pEPI) tool based on Big Dataanalytics to generate accurate risk estimates & early detect livestock health problems.The approach is to develop a pEPI set of tools based on data mining techniquesthat allows the early detection of swine health problems and generates real-time risk estimatestaking into account multi-level, space-time-genomic data and sends automatic notifications toveterinarians and producers when certain risk criteria/pre-defined thresholds are identified (i.e.,personalized early-warning system). To achieve this objective, we will first adapt severaldata mining methods to solve specific problems in the swine industry using the data describedbefore and in collaboration with pilot farms. Pilot farms will consist of 5 commercialswine operations in various Midwest states. Those swine operations have more than 1000 premisesand over 2 million pigs in total. Variables, including data from swinemanagement software, diagnostic data and precision farming devices will be obtained from thosefarms. Once evaluated, the best performing data mining methods will be selected to be integratedinto a user-friendly interface. Additional data mining methods can be explored and included infuture projects.Objective 3: Integrate all pEPI algorithms in a user-friendly online platform for their use byveterinarians and stakeholders.The approach will be to integrate the best performingdata mining methods into a user-friendly online platform to facilitate the operational use andinterpretation of the pEPI tool.