Medical multiparametric time course prognoses applied to kidney function assessments

In this paper, we describe an approach to utilize case-based reasoning meth ods for trend prognoses for the monitoring of the kidney function in an Int ensive Care Unit (ICU) setting. Since using conventional methods for reason ing over time does not fit for course predictions with poor medical knowled ge of typical course patterns, we have developed abstraction methods suitab le for integration into our case-based reasoning system ICONS. These method s combine medical experience with prognoses of multiparametric courses. On the ICU, the monitoring system NIMON provides a daily report based on curre nt measured and calculated kidney function parameters. Subsequently, we gen erate course-characteristic trend descriptions of the renal function over t he course of time. Using case-based reasoning retrieval methods, we search in the case base for courses similar to the current trend descriptions. Fin ally, we present the current course together with similar courses as compar isons and as probable prognoses to the user. We applied case-based reasonin g methods in a domain which seemed reserved for statistical methods and con ventional temporal reasoning. (C) 1999 Elsevier Science Ireland Ltd. All ri ghts reserved.