The Homer Warner Center for Medical Informatics is creating tools to automate the development process of clinical decision support applications. Our Ontology-Driven Diagnostic Modeling System (ODMS) currently applies ontologies and automated systems for analysis to create applications for diagnostic screening.
The initial ODMS is designed to automatically derive Bayesian diagnostic systems from clinical data sets. It uses medical knowledge stored in a disease-focused ontology to automate the process of identifying and extracting data from an enterprise data warehouse. The ODMS employs natural language processing tools to extract data that is not available in a coded and structured form. Once the data is extracted, an initial analysis is performed. The analysis results in the development of one or more proposed diagnostic models and the evaluation of these models with a focus on their graphical exploration.