Search research reports:
Regional Animal Health Data Warehousing and Data Mining System
C.N. Carter, W. Northington, J. Smith
Livestock Disease Diagnostic Center
The Livestock Disease Diagnostic Center at the University of Kentucky and the Breathitt Veterinary Center at Murray State University have animal health archives dating back to 1993. These collections represent over one million clinical cases that were processed through these two laboratories. A state-of-the-art, fully graphical case accessioning system has been developed and tested in a previous project that will capture all diagnostic medicine findings for current and future cases.
The intent here is to build a data warehouse that will combine the legacy case findings and the findings that will be captured by the new laboratory information management system (LIMS). The legacy data are currently residing in a proprietary, custom developed format that is not ODBC compliant. An inventory of each data element and data class must be taken for both the legacy and the new LIMS system. A new data base must then be designed that will accommodate the data elements from both systems. Finally, a means of importing the archived and current data elements into a reasonably flat, un-normalized data warehouse structure must be developed.
2010 Project Description
A laboratory information management data warehouse has been developed, configured and implemented at both the University of Kentucky Veterinary Diagnostic Laboratory and the Murray State University Breathitt Veterinary Center. Medical terminology tables have been at least 50% mapped to SNOMED concepts to allow for the fusion of data extracted by the two laboratory units. Data mining tools have been developed and tested to allow canned and ad hoc (custom) queries of the animal health data warehouse.
The overall impact of the development of the data warehouse and data mining tools is to improve animal health situational awareness throughout the state and for surrounding states. Recurring and ad hoc queries are used to better inform the Office of the Kentucky State Veterinarian and the USDA Area Veterinarian in Charge. Finally, data sets extracted from the warehouse drive the delivery of GIS disease reporting and animal health trend reporting for Kentucky animal agriculture.
Carter CN, Vanzant E, Odoi A, Smith J, Dwyer R, Riley J, Stepusin R: Supercomputer-Based Animal Health Risk Forecasting, Proceed 147th American Veterinary Medical Association, July, 2010