Challenges of unobtrusive monitoring of cardiac activity - Methods for artifact detection

Abstract:
Modern technologies for contactless measurement of cardiac activity provide the opportunity of an unobtrusive preventive medical monitoring. In contrast to a spot measurement, long-term monitoring, which covers the daily life, is possible. Thus, the real daily progress of the individual health status is reflected. However, all unobtrusive monitoring techniques involve new challenges, for which robust solutions need to be found. The two most important aspects are probably context detection and the severe motion artifacts, which may completely disrupt a measurement. In this paper, the challenges are identified and algorithms for detection of artifacts are presented and compared. The algorithms are based on a classification of detected peaks.