Sofiane Boudaoud

Combination of Canonical Correlation Analysis and Empirical Mode Decomposition Applied to Denoising the Labor Electrohysterogram

The electrohysterogram (EHG) is often corrupted by electronic and electromagnetic noise as well as movement artifacts, skeletal electromyogram, and ECGs from both mother and fetus. The interfering signals are sporadic and/or have spectra overlapping the spectra of the signals of interest rendering classical filtering ineffective. In the absence of efficient methods for denoising the monopolar EHG signal, bipolar methods are usually used. In this paper, we propose a novel combination of blind source separation using canonical correlation analysis (BSS_CCA) and empirical mode decomposition (EMD) methods to denoise monopolar EHG. We first extract the uterine bursts by using BSS_CCA then the biggest part of any residual noise is removed from the bursts by EMD. Our algorithm, called CCA_EMD, was compared with wavelet filtering and independent component analysis. We also compared CCA_EMD with the corresponding bipolar signals to demonstrate that the new method gives signals that have not been degraded by the new method. The proposed method successfully removed artifacts from the signal without altering the underlying uterine activity as observed by bipolar methods. The CCA_EMD algorithm performed considerably better than the comparison methods.

Corrected Integral Shape Averaging Applied to Obstructive Sleep Apnea Detection from the Electrocardiogram

We present a technique called corrected integral shape averaging (CISA) for quantifying shape and shape differences in a set of signals. CISA can be used to account for signal differences which are purely due to affine time warping (jitter and dilation/compression), and hence provide access to intrinsic shape fluctuations. CISA can also be used to define a distance between shapes which has useful mathematical properties; a mean shape signal for a set of signals can be defined, which minimizes the sum of squared shape distances of the set from the mean. The CISA procedure also allows joint estimation of the affine time parameters. Numerical simulations are presented to support the algorithm for obtaining the CISA mean and parameters. Since CISA provides a well-defined shape distance, it can be used in shape clustering applications based on distance measures such as-means. We present an application in which CISA shape clustering is applied to P-waves extracted from the electrocardiogram of subjects suffering from sleep apnea. The resulting shape clustering distinguishes ECG segments recorded during apnea from those recorded during normal breathing with a sensitivity of and specificity of.

Core Shape modelling of a set of curves

A new method for shape and time variability modelling of a set of curves is presented. Shape variability is captured via warping functions after time realignment of the curves. These warping functions relate normalized integrals but their meaning is different from those described in previously proposed methods for curve registration. For this purpose, a semi-parametric model, namely the Core Shape (CS) model, is proposed for shape variability characterization of a sample of curves. The curve variability is modelled as the composition of a polynomial term that accounts for time support variability and another term that accounts for intrinsic shape variability of the normalized integrals. This formalism provides specific statistical tools for shape dispersion analysis which are typically a mean shape curve, the Core Shape (CS) curve, and a shape distance, the so-called CS distance, according to the degree of specific polynomial time functions. These tools are invariant to time support variability and allow a direct access to intrinsic shape variability obtained at this polynomial degree. Also, a method for estimating shape parameters and functions of the model is presented and illustrated with simulated data. The influence of the polynomial choice is analyzed by simulation. Finally, usefulness of the proposed model for functional curve analysis is demonstrated through a real case study on Auditory Cortex Responses (ACR) analysis. A comparative study with a Curve Registration (CR) approach, namely the Self-Modelling Registration (SMR) method, is performed to better define differences in characterizing time and shape variability.

Fast generation model of high density surface EMG signals in a cylindrical conductor volume

In the course of the last decade, fast and qualitative computing power developments have undoubtedly permitted for a better and more realistic modeling of complex physiological processes. Due to this favorable environment, a fast, generic and reliable model for high density surface electromyographic (HD-sEMG) signal generation with a multilayered cylindrical description of the volume conductor is presented in this study. Its main peculiarity lies in the generation of a high resolution potential map over the skin related to active Motor Units (MUs). Indeed, the analytical calculus is fully performed in the frequency domain. HD-sEMG signals are obtained by surfacic numerical integration of the generated high resolution potential map following a variety of electrode shapes. The suggested model is implemented using parallel computing techniques as well as by using an object-oriented approach which is comprehensive enough to be fairly quickly understood, used and potentially upgraded. To illustrate the model abilities, several simulation analyses are put forward in the results section. These simulations have been performed on the same muscle anatomy while varying the number of processes in order to show significant speed improvement. Accuracy of the numerical integration method, illustrating electrode shape diversity, is also investigated in comparison to analytical transfer functions definition. An additional section provides an insight on the volume detection of a circular electrode according to its radius. Furthermore, a large scale simulation is introduced with 300MUs in the muscle and a HD-sEMG electrode grid composed of 16×16 electrodes for three constant isometric contractions in 12s. Finally, advantages and limitations of the proposed model are discussed with a focus on perspective works.

Nonlinear estimation of coupling and directionality between signals: Application to uterine EMG propagation

Understanding the direction and quantity of information flowing in a complex system is a fundamental task in signal processing. Several measures have been proposed to detect the quantity of synchronization and the directionality between time series and in physiological data. In this paper we use two methods that are widely used in synchronization and directionality analysis: Nonlinear correlation coefficient (h2) and the general synchronization (H). The performances of both methods were tested on four dimensional coupled synthetic nonlinear Rössler models. They were then applied to a single real labor contraction uterine EMG burst with the aim of using them to detect synchronization and to plot the map of direction of information flow between the whole signal channels. The results on synthetic signal show a slight superiority of H over h2. The results obtained on a single contraction are encouraging for the future use of these tools for resolving the open question of the directionality of uterine contractions and may provide a way of finding their source loci.

Surface EMG-force modelling for the biceps brachii and its experimental evaluation during isometric isotonic contractions

The aim of this study was to evaluate a surface electromyography (sEMG) signal and force model for the biceps brachii muscle during isotonic isometric contractions for an experimental set-up as well as for a simulation. The proposed model includes a new rate coding scheme and a new analytical formulation of the muscle force generation. The proposed rate coding scheme supposes varying minimum and peak firing frequencies according to motor unit (MU) type (I or II). Practically, the proposed analytical mechanogram allows us to tune the force contribution of each active MU according to its type and instantaneous firing rate. A subsequent sensitivity analysis using a Monte Carlo simulation allows deducing optimised input parameter ranges that guarantee a realistic behaviour of the proposed model according to two existing criteria and an additional one. In fact, this proposed new criterion evaluates the force generation efficiency according to neural intent. Experiments and simulations, at varying force levels and using the optimised parameter ranges, were performed to evaluate the proposed model. As a result, our study showed that the proposed sEMG–force modelling can emulate the biceps brachii behaviour during isotonic isometric contractions.

On the benefits of using HD-sEMG technique for estimating muscle force

In biomechanics modeling applications, muscle force estimators, using Hill-type model are classically fed using sEMG signals recorded from one lead per muscle. The major limitation of this methodology is the lack of muscle activation representativeness of this lead. Recently, it has emerged as an instrumental technique, based on multi-channel recordings. This technique, namely, the High-Density sEMG (HD-sEMG) technique allows us the access, at the same time, to up to 128 recording sites over a muscle. In this study, we propose to demonstrate the benefits of using such techniques in estimating more accurately the muscle force. For this purpose and using simulation, we will compare estimated force profiles using classical single bipolar and HD-sEMG techniques. Thus, we simulate HD-sEMG and bipolar signals using a recent generation model and the associated force as in Allouch et al. (2013). In fact, the simulation allows us the access to the intrinsic muscle force which is unrealizable in experiment. Next, the obtained results are discussed focusing on the benefits of using HD-sEMG for muscle force estimation.

Evaluation of HD-sEMG Probability Density Function deformations in ramp exercise.

The aim of the present study is to propose a subject-specific screening approach of High Density surface EMG (HD-sEMG) Probability Density Function (PDF) shape evolution in experimental conditions following a ramp exercise from 0% to 50% of the Maximum Voluntary Contraction (MVC) during 25 seconds of isometric contractions of the Biceps Brachii from six healthy subjects. This method uses High Order Statistics (HOS), namely the kurtosis and the skewness for PDF shape screening examined on selectively positioned Laplacian sEMG channels obtained on an 8×8 HD-sEMG grid. For each subject, the position of the Laplacian channels was chosen based on the level of muscle activation obtained from the Signal to Noise Ratio (SNR) matrix computed for the 64 sEMG signals of the grid in order to obtain independent Laplacian configurations localized in areas with high SNRs indicating high muscle activation. Afterwards, we used the Principal Component Analysis (PCA) to obtain the principal trend of the kurtosis and the skewness computed from the selected Laplacian signals according to force level variation. The obtained results show a globally common increasing HOS trend according to force increase from 0% to 50% MVC for all the subjects regardless of the anatomical, instrumental and physiological variability that usually strongly influences these trends.

Muscle force estimation using data fusion from high-density SEMG grid

The aim of the proposed work is to evaluate, by simulation, the introduction of a data fusion process from a HD-sEMG grid (8×8) to improve the muscle force estimation from sEMG signal. For this purpose, twelve electrode arrangements are combined to dimension reduction technique (PCA or channel averaging) to obtain a monodimensional sEMG signal. After, this signal is used in a sEMG-force relationship model to estimate the muscular force. In fact, two models, with different complexity, and used in the biomechanics community are studied. In the simulation, three isometric contractions are simulated (20%, 50% and 80% MVC) using a recent sEMG-force generation model. Finally, the Normalized RMS Difference (NRMSD) between the estimated force and the simulated force by the sEMG-force generation model is calculated for each combination (electrode arrangement and dimension reduction technique, force estimator). According to the obtained results, the combination PCA and Laplacian arrangement gave the best fitting using the second force estimator while the best result obtained for the first force estimator is with the Right Diagonal Bipolar (DBR) arrangement combined with channel averaging. In future works, these force estimators, combined to HD-sEMG data fusion, will be experimentally evaluated.

Study of the Muscular Force/HOS Parameters Relationship from the Surface Electromyogram

The aim of the present study is to investigate a possible relationship between High Order Statistic (HOS) parameters and muscle force. In fact, it is guessed that Motor Unit (MU) recruitment during contraction has an influence on surface EMG (sEMG) amplitude distribution shape. For this purpose, skewness and kurtosis are used to monitor variation of monopolar sEMG data according to contraction level. First, a simulation was performed to evaluate the sensitivity of both proposed parameters to physiological and instrumental parameters. Then, 3 healthy young males took part to an experimental protocol on the biceps brachii muscle. The sEMG and force data were recorded and analyzed for different voluntary contraction levels. According to the results obtained, a relationship between HOS parameters and muscle force appears to exist. However, HOS parameters are sensitive to the tested parameters.