Issue No. 01 - January/February (1998 vol. 13)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/5254.653226
Planning and calculating adequate nutritional support for sick or premature newborn infants in an intensive care unit is tedious, time-consuming work. Besides requiring considerable expert knowledge and practical experience, this task is prone to inherent, possibly fatal, calculation errors. To automate this process, we built several versions of the Vienna Expert System for Parenteral Nutrition of Neonates (VIE-PNN) knowledge-based system for prescribing parenteral (intravenous) nutrition supply. Unfortunately, clinicians did not employ these initial versions in their daily routine. This article describes our recent redesign of VIE-PNN using an HTML-based client-server architecture. We have integrated this tool into the intranet of workstations that runs a clinic's patient-data-management system. Its acceptance among physicians has been immediate. Two neonatal ICUs at the University of Vienna have put this integrated version into daily use. Factors underlying this knowledge-based system's successful operation include its ease of use, minimal required input, robustness, explanation facilities, and, most important, time savings compared to hand calculation.
A. Seyfang, S. Miksch, W. Horn and C. Popow, "Quicker, More Accurate Nutrition Plans for Newborn Infants," in IEEE Intelligent Systems, vol. 13, no. , pp. 65-69, 1998.