Aline Cabasson

Time Delay Estimation: A New Insight Into the Woody's Method

This letter introduces a new insight into the Woodys method, a well-known time delay estimator. Firstly, we show that this classical technique used to analyze a variable latency signal is suboptimal. Thereby, we present an improved version of the Woodys method that formalizes and outperforms the previous one with an unknown signal using an iterative maximum likelihood estimator. Unlike recent approaches, the problem is expressed in the time domain in order to produce a general framework that allows the inclusion of some a priori information.

Estimation and Modeling of QT-Interval Adaptation to Heart Rate Changes

This paper introduces a new method for QT-interval estimation. It consists in a batch processing mode of the improved Woodys method. Performance of this methodology is evaluated using synthetic data. In parallel, a new model of QT-interval dynamics behavior related to heart rate changes is presented. Since two kinds of QT response have been pointed out, the main idea is to split the modeling process into two steps: 1) the modeling of the fast adaptation, which is inspired by the electrical behavior at the cellular level relative to the electrical restitution curve, and 2) the modeling of the slow adaptation, inspired by experimental works at the cellular level. Both approaches are based on a low-complexity autoregressive process whose parameters are estimated using an unbiased estimator. This new modeling of QT adaptation, combined with the presented QT-estimation process, is applied to several ECG recordings with various heart rate variability dynamics. Its potential is then illustrated on ECG recorded during rest, atrial fibrillation episodes, and exercise. Meaningful results in agreement with physiological knowledge at the cellular level are obtained.

A New Modeling of the Overlapping T Wave for the efficient Estimation of the P-R Intervals during Exercise and Recovery

The analysis of the heart period series is a difficult task especially under graded exercise conditions. Having a good tool to characterize the P-R and R-R intervals, i.e. a good method of time delay estimation, would carry out a better knowledge of the neural activity during exercise and recovery in the field of pacemakers design. Unfortunately, for the P-R intervals, the problem of estimation has been rarely addressed. In this paper, we propose a new method for estimating the P-R intervals based on an iterative Maximum-Likelihood approach. The main contribution is to take into account the overlapping T wave on ECG recorded during exercise. The goal of this study is to compute a model of the T wave which overlaps the P wave and then to cancel this influence before the determination of the P-R intervals.

Comparison of sEMG processing methods during whole-body vibration exercise

The objective was to investigate the influence of surface electromyography (sEMG) processing methods on the quantification of muscle activity during whole-body vibration (WBV) exercises. sEMG activity was recorded while the participants performed squats on the platform with and without WBV. The spikes observed in the sEMG spectrum at the vibration frequency and its harmonics were deleted using state-of-the-art methods, i.e. (1) a band-stop filter, (2) a band-pass filter, and (3) spectral linear interpolation. The same filtering methods were applied on the sEMG during the no-vibration trial. The linear interpolation method showed the highest intraclass correlation coefficients (no vibration: 0.999, WBV: 0.757-0.979) with the comparison measure (unfiltered sEMG during the no-vibration trial), followed by the band-stop filter (no vibration: 0.929-0.975, WBV: 0.661-0.938). While both methods introduced a systematic bias (P < 0.001), the error increased with increasing mean values to a higher degree for the band-stop filter. After adjusting the sEMG(RMS) during WBV for the bias, the performance of the interpolation method and the band-stop filter was comparable. The band-pass filter was in poor agreement with the other methods (ICC: 0.207-0.697), unless the sEMG(RMS) was corrected for the bias (ICC ⩾ 0.931, %LOA ⩽ 32.3). In conclusion, spectral linear interpolation or a band-stop filter centered at the vibration frequency and its multiple harmonics should be applied to delete the artifacts in the sEMG signals during WBV. With the use of a band-stop filter it is recommended to correct the sEMG(RMS) for the bias as this procedure improved its performance.

Low-complexity autoregressive modeling of the fast and slow QT adaptation to heart rate changes

The aim is to develop a new model of the QT interval dynamics behavior related to heart rate changes. Since two kinds of QT response have been pointed out, the main idea is to split the modeling process into two steps: 1) the modeling of the “fast” adaptation, which is inspired by the electrical behavior at the cellular level relative to the electrical restitution curve, 2) the modeling of the “slow” adaptation, inspired by experiments works at the cellular level. Both are modeled as low-complexity autoregressive process whose parameters are computed using an unbiased estimator. The relevance of this approach is illustrated on several ECG recordings where the variations of the heart rate are various (rest, atrial fibrillation episodes, exercise). Significant results are obtained in agreement with the physiological knowledge at the cellular level.

T-P interval estimation in case of overlapping waves

Assuming two positive overlapping signals, with known shapes, the proposed method estimates the distances between their mean positions, width and area ratios. The data are two profiles representing the component shapes: no parametric model is assumed. The algorithm seeks shape equality between a linear combination of observation and first component, and the second component, in function of the area ratio. At the minimum shape difference the three parameters (distance between components, scaling factor and area ratio) are estimated. After theory, simulations are presented on Gaussian signals. Then, the method was applied on ECG signals from BSPM device during exercise on healthy people. The aim is mainly to get time distance between each T-wave and the P-wave of the following beat, on a given lead, in case of overlapping. Shape and width of the T-wave were shown to be constant before P-wave interference, which allowed taking such a real T-wave as first component model. Assumption of the same shape for the second component gave good results, as can be viewed on the reconstructed signals.