Solange Akselrod

An Efficient Algorithm for Spectral Analysis of Heart Rate Variability

We present a simple efficient algorithm for the derivation of a heart rate signal from the electrocardiogram. We demonstrate that the amplitude spectrum of this heart rate signal more closely matches that of the input signal to an integral pulse frequency modulation (IPFM) model of the hearts pacemaker than do the spectra of other ECG-derived heart rate signals. The applicability of this algorithm in cross-spectral analysis between heart rate and other physiologic signals is also discussed.

Fluctuations in autonomic nervous activity during sleep displayed by power spectrum analysis of heart rate variability.

Objective The use of an efficient noninvasive method to investigate the autonomic nervous system and cardiovascular control during sleep. Background Beat-to-beat heart rate variability displays two main components: a low-frequency (LF) one representing sympathetic and parasympathetic influence and a high-frequency (HF) component of parasympathetic origin. Sympathovagal balance can be defined as LF/HF ratio. Methods/design We reviewed normal, standardly staged all-night polysomnograms from 10 healthy children aged 6 to 17 years. Recorded 256-second traces of heart rate and respiration were sampled. Power spectra of instantaneous heart rate and respiration were computed using a fast Fourier transform method. Results The study revealed a decrease in LF during sleep, with minimal values during non-REM slow-wave sleep and elevated levels similar to those of wakefulness during REM. HF increased with sleep onset, reaching maximal values during slow-wave sleep, and behaved as a mirror image of LF. LF/HF ratio displayed changes similar to those in LF. Conclusion The sympathetic predominance that characterizes wakefulness decreases during non-REM sleep, is minimal in slow-wave sleep, and surges toward mean awake levels during REM sleep. The autonomic balance is shifted toward parasympathetic predominance during slow-wave sleep. This noninvasive method used to outline autonomic activity achieves results that are in complete agreement with those obtained with direct invasive tools.

Fluctuations in T-wave morphology and susceptibility to ventricular fibrillation

Susceptibility of the ventricles to fibrillation has been related to the degree of spatial inhomogeneity in the repolarization process. We studied the pattern of beat-to-beat fluctuations in ventricular repolarization processes in order to determine whether a relationship also exists between the temporal variability of ventricular repolarization and susceptibility to ventricular fibrillation. We used the morphology of the T-wave recorded in surface and epicardial leads as a measure of the ventricular repolarization process. The Ventricular Fibrillation Threshold (VFT) was used as the standard measure of cardiac susceptibility to fibrillation. In dog experiments, T-wave morphologic indices were computed on 1,024 sequential beats. Histogram, autocorrelation and power spectrum analyses were performed on the sequence of T-wave morphologic indices. A series of 27 experiments were performed on 20 dogs in which VFT was reduced by several different interventions--hypothermia, tachycardia and coronary artery ligation. For all three interventions we observed the same characteristic change in the pattern of T-wave morphology fluctuations. In particular, we found that as the VFT was reduced, a pattern of T-wave alternans developed. This pattern was generally not detectable by visual inspection of the ECG. It was, on the other hand, easily quantified in terms of a T-wave alternans index (TWAI) which we computed from the power spectrum of the T-wave fluctuations. In 26 of the 27 experiments, measured VFT decreased (p less than .001); in 20 of these experiments the TWAI computed from the surface ECG increased (decreased) when VFT decreased (increased) (p less than .01). In 17 experiments epicardial electrograms were recorded. In 16 of these experiments VFT decreased (p less than .001). In 16 of these 17 experiments TWAI computed from the epicardial ECG increased (decreased) when the VFT decreased (increased) (p less than .001). We conclude that statistical analysis of fluctuations in ECG complex morphology may provide a sensitive probe of ventricular vulnerability to fibrillation.

Autoimmune T cells as potential neuroprotective therapy for spinal cord injury

Autoimmune T cells against central nervous system myelin associated peptide reduce the spread of damage and promote recovery in injured rat spinal cord, findings that might lead to neuroprotective cell therapy without risk of autoimmune disease.

Spectral analysis of heart rate fluctuations. A non-invasive, sensitive method for the early diagnosis of autonomic neuropathy in diabetes mellitus.

Early detection and a quantitative evaluation of the degree of diabetic autonomic neuropathy were performed in 23 diabetic patients and 22 controls by computerized spectral analysis of beat-to-beat R-R interval variations on a continuous electrocardiogram. Simultaneous recording of cardiac and respiratory activity, R-wave detection by a fast peak detection algorithm and spectrum computation by Fast Fourier transform enabled the study of the power spectrum of heart rate fluctuations. The power of fluctuations at different frequencies is the result of sympathetic and vagal input into the sinoatrial node: this input is derived from vasomotor, baroreceptor and respiratory control loops. A marked reduction in the power of heart rate (HR) fluctuations, at all frequencies, was found in the diabetic patients as compared to controls. This indicates a depression of both parasympathetic and sympathetic activity. The difference was especially pronounced in subjects below age 65. The lowest activity was found in diabetics with concomitant peripheral neuropathy. The method described here is simple, objective, quantitative and very sensitive. It may facilitate the screening of diabetic patients for autonomic neuropathy and enable a convenient quantitative follow-up.

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.

Diffusion anisotropy MRI for quantitative assessment of recovery in injured rat spinal cord.

Spinal cord injury and its devastating consequences are the subject of intensive research aimed at reversing or at least minimizing functional loss. Research efforts focus on either attenuating the post‐injury spread of damage (secondary degeneration) or inducing some regeneration. In most of these studies, as well as in clinical situations, evaluation of the state of the injured spinal cord poses a serious difficulty. To address this problem, we carried out a diffusion‐weighted MRI experiment and developed an objective routine for quantifying anisotropy in injured rat spinal cords. Rats were subjected to a contusive injury of the spinal cord caused by a controlled weight drop. Untreated control rats were compared with rats treated with T cells specific to the central nervous system self‐antigen myelin basic protein, a form of therapy recently shown to be neuroprotective. After the rats were killed their excised spinal cords were fixed in formalin and imaged by multislice spin echo MRI, using two orthogonal diffusion gradients. Apparent diffusion coefficient (ADC) values and anisotropy ratio (AI) maps were extracted on a pixel‐by‐pixel basis. The calculated sum of AI values (SAI) for each slice was defined as a parameter representing the total amount of anisotropy. The mean‐AI and SAI values increased gradually with the distance from the site of the lesion. At the site itself, the mean‐AI and SAI values were significantly higher in the spinal cords of the treated animals than in the controls (P = 0.047, P = 0.028, respectively). These values were consistent with the score of functional locomotion. The difference was also manifested in the AI maps, which revealed well‐organized neural structure in the treated rats but not in the controls. The SAI values, AI histograms, and AI maps proved to be useful parameters for quantifying injury and recovery in an injured spinal cord. These results encourage the development of diffusion anisotropy MRI as a helpful approach for quantifying the extent of secondary degeneration and measuring recovery after spinal cord injury. Magn Reson Med 45:1–9, 2001.

Development and maturation of the autonomic nervous system in premature and full-term infants using spectral analysis of heart rate fluctuations.

ABSTRACT: The changes in the power spectra of heart rate (HR) fluctuations, in particular the total power (within 0.02–2.0 Hz) and the power in the low- (0.02–0.2 Hz) and high- (0.2–2.0 Hz) frequency ranges, were computed from the ECG and respiratory signals of 59 premature and full-term infants. The objective of the study was to investigate the development and maturation of the autonomic nervous system from the first day of extrauterine life to several weeks of postnatal age. The study population was divided into four age groups. Group A: seven 1-d-old premature infants with gestational age of 34–35 wk. Group B: 28 premature infants 7–49 d old with a conceptional age of 34–35 wk. Group C: seven 1-d-old full-term infants of 39–41 wk gestation. Group D: six premature infants 35–97 d old with a conceptional age of 39–40 wk. Mean HR (± SEM) of groups C and D combined, i.e. 135 ± 2 bpm, was significantly lower compared with groups A and B, i.e. 152 ± 2 (p < 0.01). The mean (± SEM) of the low-to high-frequency power ratio obtained from the HR power spectrum decreased progressively from 71 ± 31 in group A to 34 ± 8 in group B, 16 ± 3 in group C, and 17 ± 2 in group D. The mean low to high ratio for the combined groups C and D, 17 ± 1, was significantly lower compared with the combined group A and B, i.e. 44 ± 9 (p < 0.01). The respiratory signals showed two types of breathing patterns: a single peak in the respiratory spectrum centered around the respiratory frequency (mature type, typically found in adults); and a second type showing two separate peaks, one centered around the respiratory frequency and the other at the much lower breath amplitude modulation frequency. In both respiratory modes, the HR power spectrum usually showed a dispersed, wide pattern of the power in the high-frequency range. The progressive decline in mean HR and in low- to high-power ratio indicates a decrease in sympathovagal balance with gestational and postnatal age. This maturation might be associated with a gradual focusing into a single respiratory peak, both in the respiratory and the HR power spectrum.

Time versus frequency domain techniques for assessing baroreflex sensitivity.

Background Newer techniques to evaluate baroreflex sensitivity (BRS) are based on the analysis of blood pressure (BP) and heart rate (HR) time series in the time or frequency domain. These novel approaches are steadily gaining popularity, since they do not require injection of vasoactive substances, nor do they rely on a complex experimental set-up. Aim This review outlines and compares some basic features of the latest methods to assess spontaneous baroreflex function. Results Modern techniques for the estimation of spontaneous BRS are based on a variety of signal processing schemes and derive information on the baroreflex function from different perspectives. Thus factors such as respiration and other non-stationary agents may have different influences on the estimates provided by each of these approaches. Notwithstanding such individual specificity, however, it has been observed that in several physiological and pathophysiological conditions these techniques often provide comparable information on BRS changes over time, particularly when the estimates are averaged over time windows of a few minutes. Conclusions Due to the general agreement in the pattern of BRS among most modern methods, it seems reasonable to employ the most validated of these techniques, for which data obtained in several studies are already available.

Autonomic cardiovascular control in children with obstructive sleep apnea

Autonomic cardiorespiratory control changes with sleep-wake states and is influenced by sleep-related breathing disorders. Power spectrum (PS) analysis of instantaneous fluctuations in heart rate (HR) is used to investigate the role of the autonomic nervous system (ANS) in cardiorespiratory control. The two spectral regions of interest are the low frequency component (LF) and high frequency component (HF).The aim of the present study was to investigate the autonomic cardiorespiratory control in children with obstructive sleep apnea (OSA) syndrome. We studied 10 children with OSA versus 10 normal children. All subjects underwent whole night polysomnography. Spectral analysis of the HR and breathing signals was performed for 256 second long, artifact-free epochs in each sleep-wake state. The LF power was higher in the OSA group compared with control subjects for all states, reflecting enhanced sympathetic activity in OSA subjects. The results indicated sympathetic predominance during REM sleep in all subjects and parasympathetic predominance in slow wave sleep only in controls. The autonomic balance (LF/HF) was significantly higher in OSA patients than in control subjects, at all stages during night sleep, and while awake before sleep onset. An index of overall autonomic balance (ABI) was computed for each subject and correlated well with the measured respiratory disturbance index (RDI).