Automatic Parameter Extraction from Capacitive ECG Measurements

Capacitive ECG sensing allows acquiring an ECG, even through clothing. Due to its simple handling it is suitable for unsupervised monitoring. However, because capacitively coupled signals may differ from conventional ECGs, direct application of algorithms for parameter extraction to capacitive ECGs...

Full description

Saved in:
Bibliographic Details
Published inCardiovascular engineering and technology Vol. 3; no. 3; pp. 319 - 332
Main Authors Eilebrecht, Benjamin, Henriques, Jorge, Rocha, Teresa, Walter, Marian, Paredes, Simão, de Carvalho, Paulo, Czaplik, Michael, Leonhardt, Steffen
Format Journal Article
LanguageEnglish
Published Boston Springer US 01.09.2012
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:Capacitive ECG sensing allows acquiring an ECG, even through clothing. Due to its simple handling it is suitable for unsupervised monitoring. However, because capacitively coupled signals may differ from conventional ECGs, direct application of algorithms for parameter extraction to capacitive ECGs still needs to be analyzed. This paper analyzes the applicability of algorithms for ECG segmentation to capacitive ECG measurements. Data from a medical study with 107 subjects, where conventional and capacitive ECG were recorded simultaneously, was used for a two step parameter extraction by means of multi-scale morphological derivative transform and time-frequency analysis based algorithms. RR interval, QT interval, PQ interval and QRS duration resulting from the capacitive ECG were calculated by means of the algorithms and compared to the manually annotated data from the reference ECG. RR intervals were computed by the algorithms appropriately (mean deviation: 50 ms), calculation of QT interval, PQ interval and QRS duration yielded mean deviations of 60, 20, and 30 ms respectively. Differences of the mean value of QT duration for subjects with sinus rhythm and atrial fibrillation ( p  = 0.0414) and the QRS duration for patients with and without bundle branch block ( p  = 0.05) could be observed. An algorithmic detection of the RR intervals and the detection of heart rhythm disturbances in capacitive ECG signals are possible. Reasons for the deviation of QT interval, QT and QRS duration are deformations of the capacitive ECG signal. Their reasons should be reconsidered before unsupervised monitoring by means of capacitive ECG is possible.
ISSN:1869-408X
1869-4098
DOI:10.1007/s13239-012-0101-y