Clinical Dynamic Decision Support System based on temporal association rules

Abstract

Nosocomial Infections (NI) have been the major causes of morbidity and mortality of patients in intensive care units (ICUs) particularly in developing countries. Intensive surveillance and preventive measures is an effective element to fight against NI. Based on the temporal data recorded daily in the intensive care unit (ICU) and the help of some physicians, we plan to develop a Clinical Dynamic Decision Support System (CDDSS) based on knowledge discovery in databases (KDD) to help Physicians to predict and prevent NI. The CDDSS aims to the daily estimation of the NI occurrence probability, in the ICU patient hospitalization. The goal is to be able to anticipate if the association of some factors will support the appearance of the infections on the basis of patient histories. We propose to develop an algorithm for mining temporal association rules to extract temporal information. The discovery of temporal pattern would help them to take measures at time.

Enhanced visual data mining process for dynamic decision-making

“…The association rules mining can be considered as well used data mining technique for decision-making [31] [59]. It is the process of finding interesting and unexpected rules of the form of A=>B from temporal data.…”