Laurence Keselbrener

Selective discrete Fourier transform algorithm for time-frequency analysis: method and application on simulated and cardiovascular signals

The Selective Discrete Fourier transform (DFT) Algorithm [SDA] method for the calculation and display of time-frequency distribution has been developed and validated. For each time and frequency, the algorithm selects the shortest required trace length and calculates the corresponding spectral component by means of DFT. This approach can be extended to any cardiovascular related signal and provides time-dependent power spectra which are intuitively easy to consider, due to their close relation to the classical spectral analysis approach. The optimal parameters of the SDA for cardiovascular-like signals were chosen. The SDA perform standard spectral analysis on stationary simulated signals as well as reliably detect abrupt changes in the frequency content of nonstationary signals. The SDA applied during a stimulated respiration experiment, accurately; detected the changes in the frequency location and amplitude of the respiratory peak in the heart rate (HR) spectrum. It also detected and quantified the expected increase in vagal tone during vagal stimuli. Furthermore, the HR time-dependent power spectrum displayed the increase in sympathetic activity and the vagal withdrawal on standing. Such transient changes in HR control would have been smeared out by standard heart rate variability (HRV), which requires consideration of long trace lengths. The SDA provides a reliable tool for the evaluation and quantification of the control exerted by the Central Nervous System, during clinical and experimental procedures resulting in nonstationary signals.

Nonlinear high pass filter for R-wave detection in ECG signal.

A simple and easily implemented method for R-wave detection from ECG signals is presented. The method is based on the subtraction of a filtered version of the signal. The filter we used is a nonlinear median filter. The median filter is applied to the ECG signal. It results in a smoothed version of the signal, without any reminder of the R waves. This smoothed signal is then subtracted from the original and the resulting signal presents undistorted R-waves, without baseline drift. A simple threshold detection can then be performed on the filtered signal. Results are presented for simulated signals, with sinusoidal and step baseline drifts, as well as ECG complex shape change. The detection is accurate and the average error we obtained for 300 s length signals was of the order of 10(-8)s, for RR intervals of 1 s. Results are also presented for a real experimental signal with strong baseline drift, and the accuracy of the detection can be observed.

Is fatigue in patients with multiple sclerosis related to autonomic dysfunction

Time-dependent frequency decomposition of fluctuations in cardiovascular signals (heart rate [HR], blood pressure, and blood flow) provides noninvasive and quantitative evaluation of autonomic activity during transient and steady-state conditions. This method was applied during a change of position from supine to standing in patients with multiple sclerosis (MS) who experienced unexplained fatigue and in age-matched control subjects.No difference in response to standing, as reflected in the time domain parameters (mean HR, mean blood pressure, and mean blood flow), was observed between patients with MS and control subjects. Moreover, no difference was observed in very-low-frequency and low-frequency (related to sympathetic activity) content of HR, blood pressure, blood flow, or high-frequency content of HR (related to parasympathetic activity). The only spectral estimates that showed a significant diffenence between groups were the ratio of low-frequency to high-frequency content of HR and low-frequency content of HR normalized to total power. Both these parameters provide an estimate of the sympathovagal balance. A significant increase in these two estimates on standing was observed in control subjects only, indicating possible impairment of the sympathovagal balance response to standing in patients with MS who experienced fatigue. The authors observed a significant age dependence between close age subgroups, which occurred in the MS group only and was observed in some of the investigated spectral estimates that reflect vagal activity. Therefore, the authors assumed that age-related reduction in vagal activity occurred earlier in patients with MS who experienced fatigue. This reduction could also explain the lack of increase in the sympathovagal balance on standing. To validate this enhanced age dependence, further investigation should be performed in a larger group of subjects with a wider age range.

Autonomic response to change of posture among normal and mild-hypertensive adults: investigation by time-dependent spectral analysis.

In this study, we applied the time-dependent spectral analysis approach (SDA) to investigate the autonomic changes occurring during a transition from supine to standing position (CP), in normal and unmedicated mild hypertensive subjects. The SDA method enables an accurate follow-up of the instantaneous changes in autonomic activity, even during the unsteady phase of the transition, where sudden changes in heart rate (HR) and arterial blood pressure (ABP) are observed. We were able to quantify the vagal withdrawal (reflected in the high frequency component of the time-dependent spectrum of HR fluctuations) in the immediate response to CP and the more slowly following sympathetic increase (reflected in the low frequency component of ABP). This general pattern was observed in both groups. In addition, our results identified an altered sympathetic response to CP in mild-hypertensives, as compared to normal adults. Their basal sympathetic activity is enhanced (higher mean HR and increased low frequency fluctuations in ABP) and their response to CP is reduced, as reflected only in the LF content of ABP fluctuations, relative to normals. No difference was observed in HR fluctuations, showing that there is no parasympathetically mediated alteration of the baroreflex control of HR in mild-hypertension.

Nonlinear filters applied on computerized axial tomography : theory and phantom images

New nonlinear image processing techniques, in particular smoothing based on the understanding of the image, may create computerized tomography (CT) images of good quality using less radiation. Such techniques may be applied before the reconstruction and particularly after it. Current CT scanners use strong linear low-pass filters applied to the CT projections, reducing noise but also deteriorating the resolution of the image. The method in this study was to apply a weak low-pass filter on the projections, to perform the reconstruction, and only then to apply a nonlinear filter on the image. Various kinds of nonlinear filters were investigated based on the fact that the image is approximately piecewise constant. The filters were applied with many values of several parameters and the effects on the spatial resolution and the noise reduction were evaluated. The signal-to-noise ratio of a high-contrast phantom image processed were compared with the nonlinear filter, with the SNR of the phantom images obtained with the built-in CT linear filters in two scanning modes, the normal and the ultra high resolution modes. It was found that the nonlinear filters improve the SNR of the image, compared to the built-in filters, about three times for the normal mode and twice for the UHR scanning mode. The most successful filter on low-contrast phantom image was applied and it also seems to lead to promising results. These results seem to show that applying nonlinear filters on CT images might lead to better image quality than using the current linear filters.(ABSTRACT TRUNCATED AT 250 WORDS)

Myocardial transit time of the echocardiographic contrast media

The mean transit time of a tracer through a sample of tissue is a quantitative marker most closely related to regional tissue blood flow. Therefore, an accurate estimation of the mean time of transit of an ultrasonic tracer through a sample of myocardial tissue, obtained by contrast echocardiography, may provide a quantitative noninvasive estimate of myocardial perfusion. We hereby present an algorithm for the determination of the mean transit time by computerized analysis of a series of contrast-enhanced echocardiographic images. The algorithm comprises the evaluation of the echocardiographic impulse response function of a selected region of interest, using a deconvolution technique based on a fast Fourier transform and a frequency domain division of the videointensities measured in the sample, by that measured in a predetermined reference region. An extensive computer simulation study was designed to facilitate the optimization of the steps of analysis. We present the results of the evaluation study performed in order to assess the accuracy of the procedure in computer-simulated echocardiographic images. Within a wide range of parameters chosen to define these functions, the analysis is shown to be essentially independent of the rise and decay times of the impulse response function of the tissue sample as well as of the simulated intensities. The effects of random noise introduced into the simulated intensity curves and of their variable width were investigated. The mean transit time was found to be accurately evaluated within about 10% of error for the variety of widths and noise levels permitted. The reconvolution error did not correlate with the accuracy of the evaluation of the mean transit time, indicating that the reconvolution error cannot be used as an estimate of the accuracy of the procedure. The numerical methods and the results of the computer study are discussed in detail. The approach is proposed to be used as part of a more general technique for the quantitative measurement of regional myocardial tissue blood flow.

Estimation of autonomic response based on individually determined time axis

The analysis of the time-dependence of autonomic response requires: 1. A reliable procedure for the quantification of autonomic activity under nonsteady conditions, such as an algorithm for time-frequency decomposition (ex. SDA. Wigner-Ville, or others). 2. The choice of an adequate time scale for focusing on the data: (a) the regular, universal time scale, independent of the unsteady physiological conditions, or (b) a time axis defined by specific events related to an applied perturbation, as the indicators of specific experimental or physiological conditions, so that each individual is considered according to his own intrinsic time scale. The alignment of the various subjects according to their intrinsic time scale, reflecting their individual response mechanisms, may help to disclose a common pattern of autonomic function. Using an absolute time scale to align and average results for different subjects may obscure the underlying mechanisms. Several examples of autonomic challenges are presented, in which the use of an individual time scale contributes to unveil a typical response pattern: tilt test in vasovagal syncope, the autonomic effect of active standing on hypertension, and the autonomic response to acute hypoxia.

Time–frequency analysis of transient signals – application to cardiovascular control

A method for time–frequency decomposition (SDA) is presented for the analysis of cardiovascular signals, during steady state as well as under transient conditions. The SDA is applied to a simulated noisy non-stationary signal. It reliably discloses the time evolution of the different spectral components of the signal and does not present noise propagation as other time–frequency methods, such as Wigner–Ville distribution does. A comparison with the well-known short-time Fourier transform is also performed for non-stationary simulated signal showing that the SDA achieves a higher time–frequency resolution.

Autonomic Responses to Blockades and Provocations

Specific autonomic blockades and provocations have been widely used to identify dysfunction in autonomic control. The strength of these tests has lately been enhanced by the availability of novel quantitative approaches to determine the autonomic response, even under transient conditions. The response to autonomic manoeuvres was initially analysed in the time domain, by measures on RR intervals or blood pressure (BP). At a later stage, autonomic function was also analysed in the frequency domain.1,2,3 However, the stationarity assumption of standard spectral analysis, (based on Fourier transform [FT] or auto-regressive [AR] modelling), proscribes this approach for the evaluation of a brief or transient response to autonomic blockades or provocations. Lately, methods of time-frequency analysis, have been developed in order to overcome this stationarity limitation. Their application to cardiovascular signals has allowed obtaining information on the system, even during strongly time-dependent phases of the intervention.

Spectral analysis of left ventricular area variability as a tool to improve the understanding of cardiac autonomic control

Spectral analysis of the fluctuations in heart rate (HR) or blood pressure (BP) has been extensively used as a tool for the noninvasive assessment of autonomic control of the heart. The recently developed echocardiographic acoustic quantification allows noninvasive continuous measurement of the left ventricular cross-sectional area (LVA) signal. In this study, we investigated whether the LVA signal, and more specifically its fluctuations, can be reliably subjected to spectral analysis, and whether the results of such analysis may improve the understanding of the cardiovascular control mechanisms. Our results show that the general pattern of power spectra of LVA fluctuations, as well as their reproducibility, is similar to the power spectra of HR and BP fluctuations. Analysis of LVA signals obtained in normal subjects at rest as well as under vagal blockade and under held respiration, and in patients with known autonomic dysfunction, showed significant differences between groups and states. The effects of age, related to the reduction in parasympathetic activity, were not evident in the spectral content of the LVA and BP signals. The high frequency LVA fluctuations are mainly of mechanical origin, since they were eliminated by breath-holding. We observed an increase in the high frequency LVA fluctuations under vagal blockade, indicating that under normal (control) conditions, these high frequency fluctuations are attenuated by parasympathetic activity. The enhancement in high frequency fluctuations in LVA observed in diabetic patients can thus be attributed to reduced parasympathetic activity. The analysis of LVA variability may be used as a tool for basic research and, possibly, as a quantitative clinical measure for specific disease states.