Framework for eliciting knowledge for a medical laboratory diagnostic expert system

Medical laboratory diagnostic report system in developing countries like Nigeria is awash with data that the medical expert is expected to correctly classify or interpret visually in consultation with medical literature. The advent of more system components (diseases, tests, et cetera) and growing pressure caused mostly by shortage of medical experts render a process that is time consuming, costly and unreliable. With proper knowledge elicitation, an expert system can automate the interpretation of medical diagnostic data based on knowledge elicited from human expert classifiers.There are numerous methods for eliciting expert knowledge, all aimed at ensuring that the data collected is valid and reliable. This paper proposes a framework for implementing an elicitation process aimed at extracting valuable knowledge (in production rules) required to build a medical laboratory diagnostic expert system expected to interpret medical laboratory test reports on areas of limited domains in medicine. We have implemented a prototype expert system that operationalizes this framework for tests on urine and blood samples. Preliminary results indicate that the framework is able to facilitate effective elicitation of parameters needed for a medical laboratory diagnostic experts system.