The Use of Ensemble Empirical Mode Decomposition With Canonical Correlation Analysis as a Novel Artifact Removal Technique

Abstract

Abstract-Biosignal measurement and processing is increasingly being deployed in ambulatory situations particularly in connected health applications. Such an environment dramatically increases the likelihood of artifacts which can occlude features of interest and reduce the quality of information available in the signal. If multichannel recordings are available for a given signal source then there are currently a considerable range of methods which can suppress or in some cases remove the distorting effect of such artifacts. There are however considerably fewer techniques available if only a single channel measurement is available and yet single channel measurements are important where minimal instrumentation complexity is required. This paper describes a novel artifact removal technique for use in such a context. The technique known as ensemble empirical mode decomposition with canonical correlation analysis (EEMD-CCA) is capable of operating on single channel measurements. The EEMD technique is first used to decompose the single channel signal into a multi-dimensional signal. The CCA technique is then employed to isolate the artifact components from the underlying signal using second order statistics. The new technique is tested against the currently available Wavelet denoising and EEMD-ICA techniques using both electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) data and is shown to produce significantly improved results.

Estimation of Respiration rate from ECG Using Canonical Components Analysis and Ensemble Empirical Mode Decomposition

“…CCA is another technique which employs the Blind source separation (BSS)method for separating a number of mixed or contaminated signals [9]. CCA solves the BSS problem by forcing the sources to be maximally autocorrelated and mutually uncorrelated [6] . Given an input signal x, let y be a linear combination of neighboring samples (y(t)=x(t+1)+ x(t-1)).…”

Transient Artifact Reduction Algorithm (TARA) Based on Sparse Optimization

“…Several approaches have been developed for the suppression of artifacts in biomedical time series [2], [9], [15], [20], [21], [25], [31], [37], [38]. Some methods, such as those based on independent component analysis (ICA) or adaptive filtering, require the availability of multiple channels or reference signals.…”

“…Because the assumption is idealized, the upper bounds are too high in general (i.e., they do not guarantee convexity of ). Therefore, in the examples below, we halve these values, i.e., we set (37) In the non-convex case, we initialize the algorithm with the -norm solution to reduce the likelihood the algorithm becomes trapped in a poor local minimizer. For the -norm penalty, the initialization does not matter, due to its convexity.…”

“…When non-convex penalties are utilized, and can be set as in (37). Following the same considerations as in Section II-E, we set like for .…”

Time-Frequency Analysis of Cabin Noise Using EEMD-ICA Approaches

“…If we have adaptive decomposition of a single channel signal, and combination the IMFs into a new independent signal, thus solve the undetermined problem of blind source identification from single sampling signal. The combined techniques of EEMD-ICA have been applied in the study of some complex problems [11,12], showing the advantages of this method.…”