Modeling of Noisy Acceleration Signals From Quasi-Periodic Movements for Drift-Free Position Estimation.

Objective: We present a novel approach to drift-free position estimation from noisy acceleration signals, which often arise from quasi-periodic small-amplitude body movements. In contrast to the existing methods, this data-driven strategy is designed to properly describe time-variant harmonic structures in single-channel acceleration signals for low signal-to-noise ratios.

Methods: It comprises three processing steps: 1) short-time modeling of acceleration dynamics (instantaneous harmonic amplitudes and phases) in the analysis frame, 2) analytical integration that yields short-time position, and 3) overlap-add recombination for full-length position synthesis.

A low-rank matrix factorization approach for joint harmonic and baseline noise suppression in biopotential signals.

Background And Objectives: In this paper we propose a novel single-channel harmonic and baseline noise removal approach based on the low-rank matrix factorization theory. It aims to enhance spectrogram sparsity in order to significantly reduce the dimensionality of the underlying sources in the input data. Such a low-rank non-negative representation approach admits efficient noise removal.

Wavelet-based unsupervised learning method for electrocardiogram suppression in surface electromyograms.

We present a novel approach aimed at removing electrocardiogram (ECG) perturbation from single-channel surface electromyogram (EMG) recordings by means of unsupervised learning of wavelet-based intensity images. The general idea is to combine the suitability of certain wavelet decomposition bases which provide sparse electrocardiogram time-frequency representations, with the capacity of non-negative matrix factorization (NMF) for extracting patterns from images. In order to overcome convergence problems which often arise in NMF-related applications, we design a novel robust initialization strategy which ensures proper signal decomposition in a wide range of ECG contamination levels.

Unsupervised learning technique for surface electromyogram denoising from power line interference and baseline wander.

We present a novel approach to single-channel power line interference (PLI) and baseline wander (BW) removal from surface electromyograms (EMG). It is based on non-negative matrix factorization (NMF) using a priori knowledge about the interferences. It performs a linear decomposition of the input signal spectrogram into non-negative components, which represent the PLI, BW and EMG spectrogram estimates.

ECG-EMG separation by using enhanced non-negative matrix factorization.

We present a novel approach to single-channel ECG-EMG signal separation by means of enhanced non-negative matrix factorization (NMF). The approach is based on a linear decomposition of the input signal spectrogram in two non-negative components, which represent the ECG and EMG spectrogram estimates. As ECG and EMG have different time-frequency (TF) patterns, the decomposition is enhanced by reshaping the input mixture spectrogram in order to emphasize a sparse ECG over a noisy-like EMG.

Simultaneous powerline interference and baseline wander removal from ECG and EMG signals by sinusoidal modeling.

We present a compact approach to joint modeling of powerline interference (PLI) and baseline wonder (BW) for denoising of biopotential signals. Both PLI and BW are modeled by a set of harmonically related sinusoids modulated by low-order time polynomials. The sinusoids account on the harmonicity and mean instantaneous frequency of the PLI in the analysis window, while the polynomials capture the frequency and amplitude deviations from their nominal values and characterize the BW at the same time.

Nonstationary harmonic modeling for ECG removal in surface EMG signals.

We present a compact approach for mitigating the presence of electrocardiograms (ECG) in surface electromyographic (EMG) signals by means of time-variant harmonic modeling of the cardiac artifact. Heart rate and QRS complex variability, which often account for amplitude and frequency time variations of the ECG, are simultaneously captured by a set of third-order constant-coefficient polynomials modulating a stationary harmonic basis in the analysis window. Such a characterization allows us to significantly suppress ECG from the mixture by preserving most of the EMG signal content at low frequencies (less than 20 Hz).