Characterization and Handling of Disturbances Within Electrocardiographic Recordings of Different Origin

Résumé du livre

[ANGLÈS] A study of a method for automatic assessment of the quality of electrocardiogram (ECG) signals has been motivated by emerging applications such as the recording of the signals and transmission to mobile phones in unsupervised environments, or the measurement of vital signs in everyday scenarios without contact using capacitive technologies. These applications raise new challenges, since the quality of the signals considered is much worse than the standard signals recorded under favorable conditions. This work explores the possibility of using an ensemble of decision trees with signal features of diverse nature to automatically estimate the quality of electrocardiogram recordings of different origin. Furthermore, the classifier is tested as a channel selection method in an independent component analysis scheme for artifact removal in capacitive ECG recordings.