Use of the HELP Clinical Database to Build and Test Medical Knowledge

Abstract When building knowledge for expert systems, the analysis of patient observations and interpretations of those observations can provide a valuable source of information in addition to literature review and expert consultations. In the HELP system, a complete medical information environment, patient data is collected on a routine basis from all parts of the hospital and is made available to support knowledge base development projects. In return, the system interpretations enrich the patient database and further provide for data accuracy and validation. This work illustrates the integration of real patient data to the medical expertise and the medical literature for acquiring and validating medical knowledge.