Utilizing temporal data abstraction for data validation and therapy planning for artificially ventilated newborn infants

摘要: Medical diagnosis and therapy planning at modern intensive care units (ICUs) have been refined by the technical improvement of their equipment. However, bulk continuous data arising from complex monitoring systems in combination with discontinuously assessed numerical qualitative creates a rising information management problem neonatal ICUs (NICUs). We developed methods for validation which incorporate knowledge about point interval data, as well expected trend descriptions to arrive unified parameters (temporal abstraction). Our are based on schemata data-point transformation curve fitting express dynamics reactions different degrees parameters' abnormalities smoothing adjustment mechanisms keep stable. show applicability detecting anomalous system behavior early, recommending therapeutic actions, assessing effectiveness these actions within certain period. implemented our VIE-VENT, an open-loop knowledge-based artificially ventilated newborn infants. The usefulness approach illustrated examples VIE-VENT. Finally, we present first experiences using VIE-VENT real clinical setting.