Title:
|
Cancellation of cardiac interference in diaphragm EMG signals using an estimate of ECG reference signal
|
Author:
|
Torres Cebrián, Abel; Fiz Fernández, José Antonio; Jané Campos, Raimon
|
Other authors:
|
Universitat Politècnica de Catalunya. Departament d'Enginyeria de Sistemes, Automàtica i Informàtica Industrial; Institut de Bioenginyeria de Catalunya; Universitat Politècnica de Catalunya. BIOSPIN - Biomedical Signal Processing and Interpretation |
Abstract:
|
The analysis of the electromyographic signal of the diaphragm muscle (EMGdi) can provide important information in order to evaluate the respiratory muscular function. However, EMGdi signals are usually contaminated by the electrocardiographic (ECG) signal. An adaptive noise cancellation (ANC) based on event-synchronous cancellation can be used to reduce the ECG interference in the recorded EMGdi activity. In this paper, it is proposed an ANC scheme for cancelling the ECG interference in EMGdi signals using only the EMGdi signal (without acquiring the ECG signal). In this case the detection of the QRS complex has been performed directly in the EMGdi signal, and the ANC algorithm must be robust to false or missing QRS detections. Furthermore, an automatic criterion to select the adaptive constant of the LMS algorithm has been proposed (µ). The µ constant is selected automatically so that the canceling signal energy equals the energy of the reference signal (which is an estimation of the ECG interference present in the EMGdi signal). This approach optimizes the tradeoff between cancellation of ECG interference and attenuation of EMG component. A number of weights equivalent of a time window that contains several QRS complexes is selected in order to make the algorithm robust to QRS detection errors. |
Abstract:
|
Peer Reviewed |
Subject(s):
|
-Àrees temàtiques de la UPC::Enginyeria biomèdica::Electrònica biomèdica -Electromyography--Data processing -Adaptive Canceller -EMG -Diaphragm muscle -Electromiografia -Músculs -- Patologia |
Rights:
|
|
Document type:
|
Article - Submitted version Conference Object |
Share:
|
|