Towards artefact-free bio-impedance measurements: evaluation, identification and suppression of artefacts at multiple frequencies
Abstract
Objective: The bio-impedance (BI) of the body is a powerful parameter to assess the tissue composition and fluid content. Widespread implementation of BI measurement systems is limited due to the inability to detect, interpret and suppress frequently occurring artefacts during continuous use. A major proportion of the artefacts in BI measurements are caused at the skin electrode (SE) interface. Perturbations at the SE interface due to motion, applied pressure, temperature and posture change introduce distortions in the measured data which impede accurate interpretation for physiological monitoring. This paper investigates and accurately presents the effects of the aforementioned actions on the SE interface to develop effective algorithms for artefact identification and suppression at multiple frequencies in-field. Methods: Data collected from 20 subjects is used to analyze artefact associated features due to SE perturbations in BI measurements between 10kHz to 100kHz. Results: High variance noise, concept drifts and abrupt changes in measured BI baselines were identified as the main characteristics associated with artefacts due to SE interface perturbation. Artefacts were identified with a precision of upto 95% and a recall of upto 96% for all observed artefact labels. Conclusion: The characterization of artefacts due SE interface perturbations and high performance algorithms for artefact identification and suppression allow artefact free continuous monitoring of BI. Significance: The presented work enables continuous measurement of BI for non-invasive monitoring with wearable low power devices and permits accurate interpretation of many physiological processes, such as, edema onset, dehydration or urinary bladder filling, even in the presence of artefacts. Mehr anzeigen
Publikationsstatus
publishedExterne Links
Zeitschrift / Serie
IEEE Sensors JournalBand
Seiten / Artikelnummer
Verlag
IEEEThema
Bio-impedance; Artefact identification; Signal processing; Health monitoring; Skin electrode impedance; Wearable devices; Long-term monitoringOrganisationseinheit
09619 - Schürle-Finke, Simone / Schürle-Finke, Simone
01225 - D-ITET Zentr. f. projektbasiertes Lernen / D-ITET Center for Project-Based Learning