Abstract :
We discuss filtering procedures for robust extraction of a signal from noisy time series as they occur in monitoring of vital parameters in intensive care. An optimal monitoring of the haemodynamic system is important for any postoperative intensive care therapy. However, such time series exhibit trends, abrupt level changes and large spikes (outliers) as well as periods of relative stability. Also, the measurements are overlaid with a high level of noise. The challenge is to develop methods that allow a fast and reliable denoising of these time series. Noise and artifacts are to be separated from structural patterns of clinical relevance. Reviewing and extending recent work we present methods for robust online signal extraction and discuss their merits for preserving trends, abrupt shifts and extremes and for the removal of spikes.
