Real-time information extraction from data streams

Dr. rer. nat. Matthias Borowski

Weitere beteiligte Personen:
Prof. Dr. Roland Fried (Fakultät Statistik, TU Dortmund)
Prof. Dr. Michael Imhoff (Ruhr-Universität Bochum),
Dr. med. Evgeny Idelevich (UK Münster)

Finanzierung: Eigenmittel

Laufzeit: seit 2013

The real-time extraction of relevant information from data streams (i.e. from time series with high measurement frequency) requires fast and robust statistical methods that account for changing data conditions. One aim of this project is the development of such methods, e.g. for real-time trend estimation or for real-time surveillance of associations between data streams. Another aim is to apply the newly developed methods to particular clinical situations, e.g. to lower the rate of false positive alarms in intensive care monitoring systems, to predict the return of spontaneous circulation during resuscitation, or to quickly detect antibiotic-resistant bacteria.


Fried, R., Abbas, S., Borowski, M., Imhoff, M. (2016): "Online Analysis of Medical Time Series", submitted.

Borowski, M., Busse, D., Fried, R. (2014): "Robust online-surveillance of trend-coherence in multivariate data streams", Statistics and Computing, to appear. DOI: 10.1007/s11222-014-9462-4.

Borowski, M., Fried, R. (2013): "Online signal extraction by robust regression in moving windows with data-adaptive width selection", Statistics and Computing 24(4), 597-613.

Vorträge mit Abstracts:

Borowski, M., Kourelas, L., Bohn, A. (2015): "Real-time detection of trends in time series of carbon dioxide concentration in exhalation air", Biometrisches Kolloquium 2015, Dortmund.

Borowski, M., Rottmann, A., Wrede, C., Pemmerl, S., Imhoff, M. (2014): "Evaluation von Signalextraktionsverfahren zur Reduktion von Fehlalarmen intensivmedizinischer Überwachungssysteme", GMDS 2014, Göttingen.

Idelevich, E.A., Hoy, M., Borowski, M. et al. (2016): “Rapid phenotypic determination of carbapenem resistance in Gram-negative rods directly from positive blood cultures”, 68th Annual Meeting of the DGHM, Ulm.

Idelevich, E.A., Hoy, M., Borowski, M. et al. (2016): “Rapid growth-based detection of carbapenem-resistant Gram-negative bacteria using the real-time light-scattering method”, ASM Microbe, Boston, MA.