On the road to predictive smart alarms based on a networked operating room

In this work new algorithms are presented and discussed which extend the conventional minimum / maximum alarms in patient monitors. Two of the most important parameters indicating the patient´s condition during a surgical intervention are blood pressure and heart rate. Hence, these parameters are engaged in the assessment of anesthesia control. The computed course is used to generate foreseen minimum / maximum alarms. In future the algorithms should be used to estimate the future depth of the narcosis. For the prediction, the acquired vital signs are resampled and averaged in a pre-processing step first. Afterwards, a linear and polynomial extrapolation is performed for comparison reasons. Finally, the algorithms are discussed and compared using acquired data during standard surgeries. The developed algorithms predict the future course of heart rate and blood pressure up to five minutes and therefore can be a complement to the traditional history function of patient monitors.