Shop : Details
Shop
Details
Christoph Bernhard Hoog Antink
On Sensor Fusion for Multimodal Cardiorespiratory Signals
Sensor fusion describes the joint analysis of information from multiple sources and is, in the true sense of the word, vital to several areas. While automated sensor fusion is highly developed in many technical domains, for example navigation, it can be considered to be in an infant state in the medical realm. This work presents contributions to the field of sensor fusion for cardiorespiratory signals in three important subdomains.
In terms of signal modeling, the author developed a universal synthesizer that allows the representation of arbitrary cardiorespiratory signals. Based on a system of coupled oscillators, signals with both deterministic coupling and realistic statistical distributions could be obtained and validated for several modalities.
With respect to fusion algorithms, the author presents designed approaches for unobtrusive sensing, beat detection in medical data, and false alarm reduction in the intensive care unit. For the latter, novel features were combined with machine learning strategies and achieved promising results in an international competition.
Finally, to showcase an application scenario in the area of unobtrusive sensing, an armchair was equipped with several sensors. In a study using motion capture, the influence of motion artifacts on unobtrusive sensing modalities was analyzed and the potential of sensor fusion for monitoring vitals sings in future home monitoring scenarios was demonstrated.
In terms of signal modeling, the author developed a universal synthesizer that allows the representation of arbitrary cardiorespiratory signals. Based on a system of coupled oscillators, signals with both deterministic coupling and realistic statistical distributions could be obtained and validated for several modalities.
With respect to fusion algorithms, the author presents designed approaches for unobtrusive sensing, beat detection in medical data, and false alarm reduction in the intensive care unit. For the latter, novel features were combined with machine learning strategies and achieved promising results in an international competition.
Finally, to showcase an application scenario in the area of unobtrusive sensing, an armchair was equipped with several sensors. In a study using motion capture, the influence of motion artifacts on unobtrusive sensing modalities was analyzed and the potential of sensor fusion for monitoring vitals sings in future home monitoring scenarios was demonstrated.
Schlagwörter: Sensor Fusion; Multisensor Data Fusion; Medical Signal Processing; Unobtrusive Sensing; Machine Learning; ECG
Aachener Beiträge zur Medizintechnik
Herausgegeben von Univ.-Prof. Dr.-Ing. Dr. med. Steffen Leonhardt, Univ.-Prof. Dr.-Ing. Klaus Radermacher und Univ.-Prof. Dr. med. Dipl.-Ing. Thomas Schmitz-Rode, Aachen
Band 46
Export bibliographischer Daten
Shaker Verlag GmbHAm Langen Graben 15a52353 Düren
Mo. - Do. 8:00 Uhr bis 16:00 UhrFr. 8:00 Uhr bis 15:00 Uhr
Kontaktieren Sie uns. Wir helfen Ihnen gerne weiter.
