Management of Emergency health care conditions in Healthcare centres. A CBR approach

Abstract

Case Base Reasoning is a subfield in Artificial Intelligence (AI), this is a methodology that presents a foundation for a new technology of building intelligent computer aided diagnoses (decision support) using past and previous knowledge or solutions to a given problem so as to apply them to the current problem situation. Computer based knowledge management system aids in providing stable decision support for managing emergency treatment cases in health care centres. Management of past knowledge can expedite the process of solving emergency treatment cases that require routines or procedures for their proper management. This document describes a system that uses Case Base Reasoning as a methodology to help in treatment of anemic emergency cases during pregnancy in healthcare centres in rural Kenya that are in deficit of expert doctors or health officers. Since the treatment of the anemic cases are to be reasoned on the basis of similarity to past cases, It is very important on how to evaluate properly the degree of similarity between a target patient and past cases. Investigating the effect of similarity assessment on the performance of the proposed system we used PATDEX system and Precision/Recall computational model which are used as similarity metrics for case retrieval and case adaptation. We evaluate the system by testing it with real symptoms and signs as the query case for the diseases and examining how the performance of the system depends on the number of query cases.