Short-term analysis of heart-rate variability by adapted wavelet transforms

Rapid adjustments to autonomic control of cardiac rhythm at the onset of isometric exercise in healthy young adults

Sympathetic nervous system (SNS) and parasympathetic nervous system (PNS) influences on cardiac rhythm at the onset of exercise, a time of rapid autonomic adjustments, are clinically important areas of investigation. Continuous wavelet transform (CWT) involves time-frequency-based heart rate variability (HRV) analysis allowing investigation of autonomic influences on cardiac rhythm during short durations of exercise. Therefore, the purpose of this study was to characterize SNS and PNS influences on cardiac rhythm at the onset of isometric exercise in healthy young adults. CWT analysis was retrospectively applied to R-R interval data (electrocardiogram) previously collected from 14 healthy young adults (26 ± 2 years) who performed 30-s, one-legged, isometric, calf exercise at 70% maximal voluntary contraction (MVC; 70% MVC trial) or rested (0% MVC trial). Absolute and normalized low-frequency (aLF, nLF; 0.04-0.15 Hz) and high-frequency (aHF, nHF; 0.15-0.4 Hz) bands and LF/HF were used to analyze one 30-s baseline period and six 5-s time windows during the 30-s exercise (70% MVC) or rest (0% MVC). Statistical analysis involved two-way analysis of variance with post-hoc analysis. aHF, aLF, LF/HF, nHF, and nLF displayed a trial-time interaction (all p ≤ 0.027). In the 70% compared to the 0% MVC trial, aHF and nHF were lower after 5-30 s (all p ≤ 0.040), aLF was lower after 20-30 s (all p ≤ 0.011) and LF/HF and nLF were higher after 5-20 s (all p ≤ 0.045). These results indicate the reduction of the PNS influence on cardiac rhythm begins sooner than the augmentation of the SNS influence at the onset of isometric exercise in healthy young adults.

Methodology for the prediction of paroxysmal atrial fibrillation based on heart rate variability feature analysis

Atrial fibrillation (AF) is the most clinically diagnosed arrhythmia, as its prevalence increases with age, and its initial stage is paroxysmal atrial fibrillation (PAF). This pathology usually triggers hemodynamic disorders that can generate cerebrovascular accidents (CVA), causing morbidity and even death. The aim of this study is to predict the occurrence of PAF episodes in order to take precautions to prevent PAF episodes. The PhysioNet AFPDB prediction database was used to extract 77 heart rate variability (HRV) features using time domain, geometrical analysis, Poincaré plot, nonlinear analysis, detrended fluctuation analysis, autoregressive modeling, fast Fourier transform (FFT), Lomb-Scargle periodogram, wavelet packet transform (WPT) and bispectrum measurements. The number of features was reduced using the near-zero value, correlation, and recursive feature elimination (RFE) methods for time windows of 1, 2, 5, 10, and 30 min. Feature selection was performed using backwards selection, genetic algorithm, analysis of variance (ANOVA), and non-dominated sorting genetic algorithm (NSGA-III) methods, and then random forest, conditional random forest, k-nearest neighbor (KNN), and support vector machine (SVM) classification algorithms were applied and evaluated using 10-fold cross-validation. The proposed method achieved a precision of 93.24% with a 5-minute window and 89.21% with a 2-minute window, improving performance in predicting PAF when compared with similar studies in the literature.

Non-contact and real-time measurement of heart rate and heart rate variability using microwave reflectometry

In this paper, we present noncontact and noninvasive vital signal detection using a microwave reflectometer. Elimination of noise components due to random movement of human subjects has been the biggest issue for microwave measurement. Appropriate filtering, amplitude control of the reflectometer signal, and cross correlation among multiple reflectometers together with new algorithms have enabled motion artifact elimination, signal peak detection, and data processing for various parameters related to heart rate (HR) and heart rate variability (HRV). We focus here on the real time measurements of instantaneous HR and HRV for practical use. The evaluation by microwave reflectometry is completely noninvasive and feasible even through clothing, which is extremely effective for health maintenance in daily life as well as for preventing sudden death related to, for example, coronary heart disease and ventricular arrhythmia.

Electrocardiographic features for the measurement of drivers' mental workload

This study examines the effect of mental workload on the electrocardiogram (ECG) of participants driving the Lane Change Task (LCT). Different levels of mental workload were induced by a secondary task (n-back task) with three levels of difficulty. Subjective data showed a significant increase of the experienced workload over all three levels. An exploratory approach was chosen to extract a large number of rhythmical and morphological features from the ECG signal thereby identifying those which differentiated best between the levels of mental workload. No single rhythmical or morphological feature was able to differentiate between all three levels. A group of parameters were extracted which were at least able to discriminate between two levels. For future research, a combination of features is recommended to achieve best diagnosticity for different levels of mental workload.

Design of tree structured matched wavelet for HRV signals of menstrual cycle

An algorithm is presented for designing a new class of wavelets matched to the Heart Rate Variability (HRV) signals of the menstrual cycle. The proposed wavelets are used to find HRV variations between phases of menstrual cycle. The method finds the signal matching characteristics by minimising the shape feature error using Least Mean Square method. The proposed filter banks are used for the decomposition of the HRV signal. For reconstructing the original signal, the tree structure method is used. In this approach, decomposed sub-bands are selected based upon their energy in each sub-band. Thus, instead of using all sub-bands for reconstruction, sub-bands having high energy content are used for the reconstruction of signal. Thus, a lower number of sub-bands are required for reconstruction of the original signal which shows the effectiveness of newly created filter coefficients. Results show that proposed wavelets are able to differentiate HRV variations between phases of the menstrual cycle accurately than standard wavelets.

Cardiac parasympathetic outflow during dynamic exercise in humans estimated from power spectral analysis of P-P interval variability

NEW FINDINGS

What is the central question of this study? Should we use the high-frequency (HF) component of P-P interval as an index of cardiac parasympathetic nerve activity during moderate exercise? What is the main finding and its importance? The HF component of P-P interval variability remained even at a heart rate of 120-140 beats min(-1) and was further reduced by atropine, indicating incomplete cardiac vagal withdrawal during moderate exercise. The HF component of R-R interval is invalid as an estimate of cardiac parasympathetic outflow during moderate exercise; instead, the HF component of P-P interval variability should be used. The high-frequency (HF) component of R-R interval variability has been widely used as an indirect estimate of cardiac parasympathetic (vagal) outflow to the sino-atrial node of the heart. However, we have recently found that the variability of the R-R interval becomes much smaller during dynamic exercise than that of the P-P interval above a heart rate (HR) of ∼100 beats min(-1). We hypothesized that cardiac parasympathetic outflow during dynamic exercise with a higher intensity may be better estimated using the HF component of P-P interval variability. To test this hypothesis, the HF components of both P-P and R-R interval variability were analysed using a Wavelet transform during dynamic exercise. Twelve subjects performed ergometer exercise to increase HR from the baseline of 69 ± 3 beats min(-1) to three different levels of 100, 120 and 140 beats min(-1). We also examined the effect of atropine sulfate on the HF components in eight of the 12 subjects during exercise at an HR of 140 beats min(-1) . The HF component of P-P interval variability was significantly greater than that of R-R interval variability during exercise, especially at the HRs of 120 and 140 beats min(-1). The HF component of P-P interval variability was more reduced by atropine than that of R-R interval variability. We conclude that cardiac parasympathetic outflow to the sino-atrial node can be estimated better by the HF component of P-P interval variability during exercise and that cardiac parasympathetic nerve activity exists during moderate dynamic exercise.

Toward Capturing Momentary Changes of Heart Rate Variability by a Dynamic Analysis Method

The analysis of heart rate variability (HRV) has been performed on long-term electrocardiography (ECG) recordings (12~24 hours) and short-term recordings (2~5 minutes), which may not capture momentary change of HRV. In this study, we present a new method to analyze the momentary HRV (mHRV). The ECG recordings were segmented into a series of overlapped HRV analysis windows with a window length of 5 minutes and different time increments. The performance of the proposed method in delineating the dynamics of momentary HRV measurement was evaluated with four commonly used time courses of HRV measures on both synthetic time series and real ECG recordings from human subjects and dogs. Our results showed that a smaller time increment could capture more dynamical information on transient changes. Considering a too short increment such as 10 s would cause the indented time courses of the four measures, a 1-min time increment (4-min overlapping) was suggested in the analysis of mHRV in the study. ECG recordings from human subjects and dogs were used to further assess the effectiveness of the proposed method. The pilot study demonstrated that the proposed analysis of mHRV could provide more accurate assessment of the dynamical changes in cardiac activity than the conventional measures of HRV (without time overlapping). The proposed method may provide an efficient means in delineating the dynamics of momentary HRV and it would be worthy performing more investigations.

Investigating the performance improvement of HRV Indices in CHF using feature selection methods based on backward elimination and statistical significance

In this study, the best combination of short-term heart rate variability (HRV) measures was investigated to distinguish 29 patients with congestive heart failure from 54 healthy subjects in the control group. In the analysis performed, wavelet packet transform based frequency-domain measures and several non-linear parameters were used in addition to standard HRV measures. The backward elimination and unpaired statistical analysis methods were used to select the best one among all possible combinations of these measures. Five distinct typical classifiers with different parameters were evaluated in discriminating these two groups using the leave-one-out cross validation method. Each algorithm was tested 30 times to determine the repeatability of the results. The results imply that the backward elimination method gives better performance when compared to the statistical significance method in the feature selection stage. The best performance (82.75%, 96.29%, and 91.56% for the sensitivity, specificity, and accuracy) was obtained by using the SVM classifier with 27 selected features including non-linear and wavelet-based measures.

Investigation of EEG abnormalities in the early stage of Parkinson's disease

The objective of the present study was to investigate brain activity abnormalities in the early stage of Parkinson's disease (PD). To achieve this goal, eyes-closed resting state electroencephalography (EEG) signals were recorded from 15 early-stage PD patients and 15 age-matched healthy controls. The AR Burg method and the wavelet packet entropy (WPE) method were used to characterize EEG signals in different frequency bands between the groups, respectively. In the case of the AR Burg method, an increase of relative powers in the δ- and θ-band, and a decrease of relative powers in the α- and β-band were observed for patients compared with controls. For the WPE method, EEG signals from patients showed significant higher entropy over the global frequency domain. Furthermore, WPE in the γ-band of patients was higher than that of controls, while WPE in the δ-, θ-, α- and β-band were all lower. All of these changes in EEG dynamics may represent early signs of cortical dysfunction, which have potential use as biomarkers of PD in the early stage. Our findings may be further used for early intervention and early diagnosis of PD.

The study on Chan-meditation electrocardiogram by pattern analysis of continuous wavelet transform-coefficient map

OBJECTIVE: To investigate the effect of Chan-meditation on electrocardiogram (ECG) regulated by various respiratory rhythms. METHODS: ECG complexes were firstly transformed into continuous wavelet transform (CWT) coefficient map. Three schemes were employed in the interpretation of CWT map: the moment-invariants analysis, the correlation-coefficient analysis of eigenvector derived by singular value decomposition (SVD), and the analysis of variance (ANOVA). This study involved 17 subjects: 8 experimental subjects with Chan-meditation experience and 9 without any Chan-mediation experience as the control subjects in the same age range. RESULTS: The results of all 3 different schemes for interpreting the CWT map coincidently demonstrated the extraordinary state of cardiorespiratory interaction behavior for the experimental subjects breathing at higher respiratory rate. According to the ANOVA analysis, the control group exhibited statistically significant difference in CWT map of ECG complex at low respiratory rates, while experimental group did not. CONCLUSION: This might preliminarily suggest that, with slow respiration, Chan-meditation practitioners had their cardiac operation more stable than normal people.

Heart rate variability dynamics during low-dose propofol and dexmedetomidine anesthesia

Heart rate variability (HRV) has been observed to decrease during anesthesia, but changes in HRV during loss and recovery of consciousness have not been studied in detail. In this study, HRV dynamics during low-dose propofol (N = 10) and dexmedetomidine (N = 9) anesthesia were estimated by using time-varying methods. Standard time-domain and frequency-domain measures of HRV were included in the analysis. Frequency-domain parameters like low frequency (LF) and high frequency (HF) component powers were extracted from time-varying spectrum estimates obtained with a Kalman smoother algorithm. The Kalman smoother is a parametric spectrum estimation approach based on time-varying autoregressive (AR) modeling. Prior to loss of consciousness, an increase in HF component power indicating increase in vagal control of heart rate (HR) was observed for both anesthetics. The relative increase of vagal control over sympathetic control of HR was overall larger for dexmedetomidine which is in line with the known sympatholytic effect of this anesthetic. Even though the inter-individual variability in the HRV parameters was substantial, the results suggest the usefulness of HRV analysis in monitoring dexmedetomidine anesthesia.

A continuous wavelet transform-based method for time-frequency analysis of artefact-corrected heart rate variability data

Time-frequency analysis of heart rate variability (HRV) provides relevant clinical information. However, time-frequency analysis is very sensitive to artefacts. Artefacts that are present in heart rate recordings may be corrected, but this reduces the variability in the signal and therefore adversely affects the accuracy of calculated spectral estimates. To overcome this limitation of traditional techniques for time-frequency analysis, a new continuous wavelet transform (CWT)-based method was developed in which parts of the scalogram that have been affected by artefact correction are excluded from power calculations. The method was evaluated by simulating artefact correction on HRV data that were originally free of artefacts. Commonly used spectral HRV parameters were calculated by the developed method and by the short-time Fourier transform (STFT), which was used as a reference. Except for the powers in the very low-frequency and low-frequency (LF) bands, powers calculated by the STFT proved to be extremely sensitive to artefact correction. The CWT-based calculations in the high-frequency and very high-frequency bands corresponded well with their theoretical values. The standard deviations of these powers, however, increase with the number of corrected artefacts which is the result of the non-stationarity of the R-R interval series that were analysed. The powers calculated in the LF band turned out to be slightly sensitive to artefact correction, but the results were acceptable up to 20% artefact correction. Therefore, the CWT-based method provides a valuable alternative for the analysis of HRV data that cannot be guaranteed to be free of artefacts.

Measurement of heart rate variability and stress evaluation by using microwave reflectometric vital signal sensing

In this paper, we present two robust signal processing techniques for stress evaluation using a microwave reflectometric cardiopulmonary sensing instrument. These techniques enable the heart rate variability (HRV) to be recovered from measurements of body-surface dynamic motion, which is subsequently used for the stress evaluation. Specifically, two novel elements are introduced: one is a reconfiguration of the HRV from the cross-correlation function between a measurement signal and a template signal which is constructed by averaging periodic component over a measurement time. The other is a reconstruction of the HRV from the time variation of the heartbeat frequency; this is evaluated by a repetition of the maximum entropy method. These two signal processing techniques accomplish the reconstruction of the HRV, though they are completely different algorithms. For validations of our model, an experimental setup is presented and several sets of experimental data are analyzed using the two proposed signal processing techniques, which are subsequently used for the stress evaluation. The results presented herein are consistent with electrocardiogram data.

Combining predictive capabilities of transcranial doppler with electrocardiogram to predict hemorrhagic shock

Hemorrhagic shock (HS) potentially impacts the chance of survival in most traumatic injuries. Thus, it is highly desirable to maximize the survival rate in cases of blood loss by predicting the occurrence of hemorrhagic shock with biomedical signals. Since analyzing one physiological signal may not enough to accurately predict blood loss severity, two types of physiological signals - Electrocardiography (ECG) and Transcranial Doppler (TCD) - are used to discover the degree of severity. In this study, these degrees are classified as mild, moderate and severe, and also severe and non-severe. The data for this study were generated using the human simulated model of hemorrhage, which is called lower body negative pressure (LBNP). The analysis is done by applying discrete wavelet transformation (DWT). The wavelet-based features are defined using the detail and approximate coefficients and machine learning algorithms are used for classification. The objective of this study is to evaluate the improvement when analyzing ECG and TCD physiological signals together to classify the severity of blood loss. The results of this study show a prediction accuracy of 85.9% achieved by support vector machine in identifying severe/non-severe states.

Heart rate normalization in the analysis of heart rate variability in congestive heart failure

In this study, the effects of heart rate (HR) normalization in the analysis of the heart rate variability (HRV) were investigated to distinguish 29 patients with congestive heart failure from 54 healthy subjects in the control group. In the analysis performed, the best accuracy performances of optimal combination of standard short-term HRV measures and of HR-normalized short-term HRV measures are compared. A genetic algorithm is used to select the best features from among all possible combinations of these measures. A k-nearest-neighbour (KNN) classifier is used to evaluate the performances of the feature combinations in classifying these two data groups. The results imply that using both min—max and HR normalization improves the performance of the classification. The maximum accuracy is achieved as 93.98 per cent using k = 3 and k = 5 for the KNN classifier with the perfect positive predictivity values.

Control of heart rate variability by cardiac parasympathetic nerve activity during voluntary static exercise in humans with tetraplegia

Heart rate (HR) is controlled solely by via cardiac parasympathetic outflow in tetraplegic individuals, who lack supraspinal control of sympathetic outflows and circulating catecholamines but have intact vagal pathways. A high-frequency component (HF; at 0.15–0.40 Hz) of the power spectrum of HR variability and its relative value against total power (HF/Total) were assessed using a wavelet transform to identify cardiac parasympathetic outflow. The relative contribution of cardiac parasympathetic and sympathetic outflows to controlling HR was estimated by comparing the HF/Total-HR relationship between age-matched tetraplegic and normal men. Six tetraplegic men with complete cervical spinal cord injury performed static arm exercise at 35% of the maximal voluntary contraction until exhaustion. Although resting cardiac output and arterial blood pressure were lower in tetraplegic than normal subjects, HR, HF, and HF/Total were not statistically different between the two groups. When tetraplegic subjects developed the same force during exercise as normal subjects, HF and HF/Total decreased to 67–90% of the preexercise control and gradually recovered 1.5 min after exercise. The amount and time course of the changes in HF/Total during and after exercise coincided well between both groups. In contrast, the increase in HR at the start of exercise was blunted in tetraplegic compared with normal subjects, and the HR recovery following exercise was also delayed. It is likely that, although the withdrawal response of cardiac parasympathetic outflow is preserved in tetraplegic subjects, sympathetic decentralization impairs the rapid acceleration of HR at the onset of exercise and the rapid deceleration following exercise.

Combining classical HRV indices with wavelet entropy measures improves to performance in diagnosing congestive heart failure

In this study, best combination of short-term heart rate variability (HRV) measures are sought for to distinguish 29 patients with congestive heart failure (CHF) from 54 healthy subjects in the control group. In the analysis performed, in addition to the standard HRV measures, wavelet entropy measures are also used. A genetic algorithm is used to select the best ones from among all possible combinations of these measures. A k-nearest neighbor classifier is used to evaluate the performance of the feature combinations in classifying these two groups. The results imply that two combinations of all HRV measures, both of which include wavelet entropy measures, have the highest discrimination power in terms of sensitivity and specificity values.

Comparison of time and frequency domain measures of RSA in ambulatory recordings

The extent to which various measures of ambulatory respiratory sinus arrhythmia (RSA) capture the same information across conditions in different subjects remains unclear. In this study the root mean square of successive differences (RMSSD), peak valley RSA (pvRSA), and high frequency power (HF power) were assessed during ambulatory recording in 84 subjects, of which 64 were retested after about 3 years. We used covariance structure modeling to test the equality of the correlations among three RSA measures over two test days and three conditions (daytime sitting or walking and nighttime sleep) and in groups with low, medium, and high mean heart rate (HR), or low, medium, and high mean respiration rate (RR). Results showed that ambulatory RMSSD, pvRSA, and HF power are highly correlated and that their correlation is stable across time, ambulatory conditions, and a wide range of resting HR and RR values. RMSSD appears to be the most cost-efficient measure of RSA.

Correlates of the shift in heart rate variability with postures and walking by time-frequency analysis

Heart rate (HR) variability derived from electrocardiogram (ECG) can be used to assess the function of the autonomic nervous system. HR exhibits various characteristics during different physical activities attributed to the altered autonomic mediation, where it is also beneficial to reveal the autonomic shift in response to physical-activity change. In this paper, the physical-activity-related HR behaviors were delineated using a portable ECG and body acceleration recorder based on a personal digital assistant and the smoothed pseudo Wigner-Ville distribution. The results based upon eighteen subjects performing four sequential 5-min physical activities (supine, sitting, standing and spontaneous walking) showed that the high-frequency heartbeat fluctuations during supine and sitting were significantly larger than during standing, and that the ratio of low- to high-frequency fluctuation during standing was significantly higher than during supine and sitting. This could be linked with the parasympathetic predominance during supine and sitting, and a shift to sympathetic dominance while standing. During spontaneous walking, the high-frequency fluctuation was significant lower than during supine. The low- to high-frequency ratio decreased significantly from standing to spontaneous walking, which may imply an increased vagal predominance (autonomic effect) or an increased respiratory activity (mechanical effect).

Time-varying analysis of heart rate variability with kalman smoother algorithm

A time-varying parametric spectrum estimation method for analyzing nonstationary heart rate variability signals is presented. In the method, the nonstationary signal is first modeled with time-varying autoregressive model and the model parameters are estimated recursively with a Kalman smoother algorithm. The spectrum estimates for each time are then obtained from the estimated model parameters. Statistics of the obtained spectrum estimates are derived using the error propagation principle. The obtained spectrum estimates can further be decomposed into separate components and, thus, the time-variation of low and high frequency components of heart rate variability can be examined separately.

An efficient algorithm for R-R intervals series filtering

Spectral analysis of heart rate variability (HRV) constitute a simple and non invasive way to study the autonomic nervous system (ANS) activity. On-line implementation of this technique would allow to follow the evolution of the ANS activity and to track transient events during medical procedures. However, continuous spectral analysis of HRV is not reliable enough due to the difficulty to obtain a noiseless ECG signal during a long period. Indeed, the consequential effects of each ECG signal perturbation on the R-R intervals gives an erroneous evaluation of HRV spectral analysis. In this article, we describe a real time filtering algorithm for R-R intervals series. This filter is able to detect each disturbed area and to replace the erroneous samples with the most probable ones. Therefore, this method allows detecting and replacing up to 90 % of R-R series erroneous samples while keeping the real recording time and without having any effect, beyond measure, on the frequency analysis result.

Wavelet transform: A better approach for the evaluation of instantaneous changes in heart rate variability

The aim of our study was to validate the Vaidyanathan wavelet tool for HRV analysis during orthostatic testing. Two groups of normotensive male subjects were studied: 13 adolescents and 27 young adults. Both groups consisted of subjects with negative, (N-), and with positive family history for hypertension, (N+). These subjects underwent 5-minute active standing upright, preceded and followed by 5-minute periods in supine position. Continuous electrocardiogram (ECG) was recorded and HRV indices were calculated using wavelet (WT) and fast Fourier transform (FFT) simultaneously. WT and FFT data showed high level of correlation (>0.9). Due to its inherent properties, WT proved to be more informative than FFT in the analysis of the non-stationary ECG signal during orthostatic testing. WT revealed HRV dynamics more accurately since it allowed HRV evaluation for shorter intervals (60 s) than FFT (256 s). During the initial and recovery period lower parasympathetic activity (P < 0.0001; P < 0.02) and higher ratio of autonomic balance (sympathetic vs. parasympathetic) (P < 0.0001; P < 0.02) were evidenced in (N+) as compared to (N-). The upright posture was accompanied by a prompt decrease in HRV and by an elevation of the index of autonomic balance. These alterations were more pronounced in N(+). In conclusion, we believe that wavelet analysis is an appropriate approach for the estimation of HRV dynamics in non-stationary conditions. Furthermore, we demonstrate certain essential alterations in the autonomic modulation of the cardiovascular system in young normotensives with positive family history for essential hypertension.

Use of wavelet and fast Fourier transforms in pharmacodynamics

Progress has been made in the development and application of mechanism-based pharmacodynamic models for describing the drug-specific and physiological factors influencing the time course of responses to the diverse actions of drugs. However, the biological variability in biosignals and the complexity of pharmacological systems often complicate or preclude the direct application of traditional structural and nonstructural models. Mathematical transforms may be used to provide measures of drug effects, identify structural and temporal patterns, and visualize multidimensional data from analyses of biomedical signals and images. Fast Fourier transform (FFT) and wavelet analyses are two methodologies that have proven to be useful in this context. FFT converts a signal from the time domain to the frequency domain, whereas wavelet transforms colocalize in both domains and may be utilized effectively for nonstationary signals. Nonstationary drug effects are common but have not been well analyzed and characterized by other methods. In this review, we discuss specific applications of these transforms in pharmacodynamics and their potential role in ascertaining the dynamics of spatiotemporal properties of complex pharmacological systems.

Wavelet transform for analysis of heart rate variability preceding ventricular arrhythmias in patients with ischemic heart disease

INTRODUCTION

Studies evaluating changes in HRV preceding the onset of ventricular arrhythmias using conventional techniques have shown inconsistent results. Time-frequency analysis of HRV is traditionally performed using short-term Fourier transform (STFT). Wavelet transform (WT) may however be better suited for analyzing non-stationary signals such as heart rate recordings.

METHODS AND RESULTS

We studied patients with a history of myocardial infarction implanted with a defibrillator with an extended memory. The RR intervals during the 51 min preceding ventricular events requiring electrical therapy were retrieved, and HRV studied by WT and STFT. 111 episodes of ventricular arrhythmia were retrieved from 41 patients (38 males, age 64 +/- 8 years). Heart rate increased significantly before arrhythmia. There was no significant variation in low frequency / high frequency components (LF/HF) observed for the group as a whole, probably due to a great degree of heterogeneity amongst individuals. A subset of 30 patients also had heart rate recordings performed during normal ICD follow-up. WT did not show any difference in HRV before arrhythmia onset and during control conditions.

CONCLUSION

Variations in HRV before onset of ventricular arrhythmias were not apparent in this large dataset, despite use of optimal tools for studying time-frequency analysis.

Time-varying analysis of heart rate variability signals with a Kalman smoother algorithm

A time-varying parametric spectrum estimation method for analysing non-stationary heart rate variability signals is presented. As a case study, the dynamics of heart rate variability during an orthostatic test is examined. In this method, the non-stationary signal is first modelled with a time-varying autoregressive model and the model parameters are estimated recursively with a Kalman smoother algorithm. The benefit of using the Kalman smoother is that the lag error present in a Kalman filter, as well as in all other adaptive filters, can be avoided. The spectrum estimates for each time instant are then obtained from the estimated model parameters. Statistics of the obtained spectrum estimates are derived using the error propagation principle. The obtained spectrum estimates can further be decomposed into separate components and, thus, the time variation of low- and high-frequency components of heart rate variability can be examined separately. By using the presented method, high resolution time-varying spectrum estimates with no lag error can be produced. Other benefits of the method are the straightforward procedure for evaluating the statistics of the spectrum estimates and the option of spectral decomposition.

Wavelet transforms and the ECG: a review

The wavelet transform has emerged over recent years as a powerful time-frequency analysis and signal coding tool favoured for the interrogation of complex nonstationary signals. Its application to biosignal processing has been at the forefront of these developments where it has been found particularly useful in the study of these, often problematic, signals: none more so than the ECG. In this review, the emerging role of the wavelet transform in the interrogation of the ECG is discussed in detail, where both the continuous and the discrete transform are considered in turn.

Customized spectral band analysis compared with conventional Fourier analysis of heart rate variability in neonates

A customized filtering technique is introduced and compared with fast Fourier transformation (FFT) for analyzing heart rate variability (HRV) in neonates from short-term recordings. FFT is classically the most commonly used spectral technique to investigate cardiovascular fluctuations. FFT requires stability of the physiological signal within a 300 s time window that is usually analyzed in adults. Preterm infants, however, show characteristics of rapidly fluctuating heart rate and blood pressure due to an immature autonomic regulation, resulting in non-stationarity of these signals. Therefore neonatal studies use (half-overlapping or moving) windows of 64 s length within a recording time of 2-5 min. The proposed filtering technique performs a filtering operation in the frequency range of interest before calculating the spectrum, which allows it to perform an analysis of shorter periods of only 42 s. The frequency bands of interest are 0.04-0.15 Hz (low frequency, LF) and 0.4-1.5 Hz (high frequency, HF). Although conventional FFT analysis as well as the proposed alternative technique result in errors in the estimation of LF power, due to spectral leakage from the very low frequencies, FFT analysis is more sensitive to this effect. The response times show comparable behavior for both the techniques. Applying both the methods to heart rate data obtained from a neonate before and after atropine administration (inducing a wide range of HRV), shows a very significant correlation between the two methods in estimating LF and HF power. We conclude that a customized filtering technique might be beneficial for analyzing HRV in neonates because it reduces the necessary time window for signal stability.

Forecasting epilepsy from the heart rate signal

Information contained in the R-R interval series, specific to the pre-ictal period, was sought by applying an unsupervised fuzzy clustering algorithm to the N-dimensional phase space of N consecutive interval durations or the absolute value of duration differences. Data sources were individual, complex partial seizures of temporal-lobe epileptics and generalised seizures of rats rendered epileptic with hyperbaric oxygen. Forecasting success was 86% and 82% (zero false positives in resistant rats), respectively, at times ranging from 10 min to 30 s prior to seizure onset Although certain forecasting clusters predominated in the patient group and different ones predominated in the animal group, forecasting on the whole was seizure-specific. The high prediction sensitivity of this method, which matches that of EEG-based methods, seems promising. It is believed that an on-line version of the algorithm, trained on each patient's peri-ictal ECG, could serve as a basis for a simple seizure alarm system.

Wavelet-based analysis of heart-rate-dependent ECG features

BACKGROUND

Wavelet-based methods of analyzing ECG signals have been used to identify specific features in cardiac arrhythmias. Since some of these features are rate dependent, it is a requirement that they are examined across a range of physiological heart rates. The wavelet transform is a signal analysis tool that can elucidate spectral and temporal information simultaneously from complex signals, including the ECG. The aim of this study was to identify the local frequency characteristics of the ECG using a real-time wavelet scalogram and to study the rate dependence of these features.

METHODS

We examined the spectral temporal behavior of the local characteristics of the electrocardiogram (ECG) of 10 patients, in whom precise control of heart rate was achieved using right atrial pacing. Temporary reprogramming was used to adjust the paced atrial rate to predetermined values so that a rate-controlled rhythm was produced that closely resembled sinus rhythm.

RESULTS

Rate-dependent features are seen on time-frequency scalograms. As the rate increases, the temporal spacing of features decrease and the frequency bands shift upward on the plot. Two patients with abnormal atrioventricular conduction demonstrate features of Wenckebach conduction and fusion.

CONCLUSIONS

Characterization of the rate-dependent features of the ECG in a paced atrial rhythm by wavelet transform techniques has revealed some additional information not readily seen on single lead ECG analysis. This model provides a surrogate for changes that might be expected during rate changes in physiological sinus rhythm. It is envisaged that this method will offer advantages in detecting features of clinical significance that may not be readily seen by existing methods.

Heart rate variability during sedentary work and sleep in normal and sleep-deprived states

The possibility of using heart rate variability (HRV) as an indicator of sleepiness was investigated by analysing heart rate (HR) activity and electroencephalography (EEG) recordings from 10 individuals who performed a monotonous attention task for 120 min in both sleep-deprived and rested states. In both conditions, measurements were collected during 60 min of sleep immediately following a 120 min of non-sleep (awake phase). Although HR decreased significantly in both the rested and the sleep-deprived states during the awake phase, HR significantly changed sooner when subjects were sleep-deprived than when they were rested. No significant changes in HRV were found during the awake phase; however, HRV correlated significantly with alpha and theta power densities when rested but not when sleep-deprived. During the sleep phase, the total HRV and very low and low frequency HRV components significantly decreased approximately 40 min after sleeping in the sleep-deprived condition. These HRV components were also significantly and negatively correlated with delta power densities. HRV does not seem to be a viable indicator of sleepiness; however, HRV may be useful for determining sleep stages.

Wavelet packet transform for R-R interval variability

INTRODUCTION

Wavelet transform is used for time-frequency analysis. Recently, discrete wavelet transform (DWT) has been used to analyze R-R interval or heart rate variability. However, we hypothesized that wavelet packet transform (WPT) is a better way to analyze such variability. In the present study, we compared resolution of frequency band and amplitude, which are used for analysis of the variability, with DWT and WPT, followed by Hilbert transform.

METHODS

A chirp signal which covers all frequency bands used for R-R interval variability was employed as a simulated signal. Levels 1-6 of DWT and level 3 of WPT were used for signal analysis. Amplitudes of the gained signal were evaluated with Hilbert transform. Differences in error of the gained amplitude from expected amplitude between CWT and DWT for low-frequency (LF) and high-frequency (HF) components were compared. To evaluate time-dependent changes in R-R interval variability, head-up tilt (HUT) was employed as an orthostatic challenge.

RESULTS

Errors for both HF and LF, derived from the simulated signal with WPT, were significantly smaller than those of DWT. With HUT, time dependent changes in LF, HF, and LF/HF were observed.

DISCUSSION

Although DWT is a valuable method for time-frequency analysis, WPT is a more appropriate method to utilize wavelet transform due to the equivalent resolution of the gained frequency band. WPT for time-frequency analysis improves analysis of time-dependent changes in R-R interval variability.

Wavelet transform analysis of heart rate variability to assess the autonomic changes associated with spontaneous coronary spasm of variant angina

We used Wavelet transform (WT) to investigate whether variation in autonomic tone was associated with spontaneous coronary spasm in patients with variant angina by analysis of heart rate variability (HRV). Twenty-one episodes preceding ST-segment elevation were selected under Holter monitoring in 12 men and 3 women with variant angina. HRV indices were calculated at 10 second intervals with the continuous WT, and analyzed within 30 minutes preceding ST-segment elevation. High frequency (HF; 0.15 approximately 2.00 Hz) increased significantly during the 4 minutes prior to ST-segment elevation, low frequency (LF; 0.04 approximately 0.15 Hz) decreased significantly during the period from 10 to 5 minutes and increased significantly during the 2 minutes prior to ST-segment elevation, the LF/HF ratio decreased significantly during the period from 10 to 3 minutes and increased significantly during the 2 minutes prior to ST-segment elevation. The RR interval decreased significantly during the 2 minutes prior to ST-segment elevation. These results suggest that the acute variation in autonomic tone was associated with spontaneous coronary spasm in patients with variant angina. A reduction in sympathetic activity, then enhancement of vagal activity may play a key role in triggering the spontaneous coronary spasm, and the secondary activation of sympathetic activity may worsen the coronary spasm resulting in the attack.

Combined index of heart rate variability and oximetry in screening for the sleep apnoea/hypopnoea syndrome

Many sleep centres employ a preliminary screening test in order to reduce the number of polysomnographies required in the routine diagnosis of the sleep apnoea/hypopnoea syndrome (SAHS). We investigated the combination of heart rate and oximetry information as a means of performing this test. A retrospective study of 100 patients with suspected SAHS was made. All patients had in-hospital polysomnography on one night. We estimated the number of respiratory event-related arousals by counting the number of autonomic arousals (assessed on the basis of changes in the heart interbeat interval) that were coincident with a rise in oximetry. The hourly index of such events was denoted the "cardiac-oximetry disturbance index" (CODI). The median apnoea/hypopnoea index (AHI) was 16.5 (range 1.0-93.6) h-1. The CODI correlated significantly with the AHI (Spearman correlation coefficient rs = 0.88, P < 0.01), and the area (+/- standard error) under the receiver operating characteristic (ROC) was 0.94 +/- 0.05. Oximetry alone (based on 4% dips) was a less effective screening test (rs = 0.80, P < 0.01; area under ROC 0.83 +/- 0.06). Using 2% dips in oximetry offered comparable performance with the CODI (rs = 0.91, P < 0.01; area under ROC 0.93 +/- 0.04). The CODI was better correlated with the electroencephalograph arousal index (rs = 0.84, P < 0.01) than was oximetry (2% dips, rs = 0.57, P < 0.01). The CODI algorithm also offers an informal measure of self-validation: a large discrepancy between the number of autonomic arousals and the number of rises in oximetry indicates the presence of autonomic arousals without changes in oximetry (or vice versa). This self-validation mechanism identified several patients in this study, and may be useful in identifying sleep disruption due to chronic pain or other causes.

Wavelet decomposition of cardiovascular signals for baroreceptor function tests in pigs

In this paper, the discrete wavelet transform (DWT) was applied to analyze the fluctuations in RR interval and systolic arterial pressure (SAP) recorded from eight alpha-chloralose anesthetized pigs. Our aim was to characterize the autonomic modulation before and after cardiac autonomic blockade and during baroreflex function tests. The instantaneous power of decomposed low-frequency (LF) and high-frequency (HF) components was used for a time-variant spectral analysis. Our results suggested that transient events and changes in autonomic modulation were detected with high temporal resolution. A nonlinear relationship between RR interval and SAP during pharmacologically induced changes in blood pressure was found, when the superimposed effect of respiratory sinus arrhythmia was removed. In addition, the baroslopes were nearly linear when both the LF and HF components were removed using DWT decomposition.

Tracking rhythmicity in nonstationary quasi-periodic biomedical signals using adaptive time-varying covariance

A time-varying covariance method for detecting and quantifying the evolution of rhythmicity (frequency) in persistently varying quasi-periodic nonstationary signals is presented. The basic method, evaluated using chirp signals, utilizes a shifting window of fixed length. A substantial reduction in estimation bias and variability are obtained by utilizing an adaptive window whose length is dependent on past frequency estimates. The adaptive window yields estimates that are comparable in accuracy to those obtained using high-resolution time-frequency representation but with lower computation requirements and the potential for on-line application. Finally, an example of the application of the method for analyzing a neural recording is also illustrated.

Wavelet transform analysis of heart rate variability during myocardial ischaemia

Analysis of heart rate variability (HRV) is a valuable, non-invasive method for quantifying autonomic cardiac control in humans. Frequency-domain analysis of HRV involving myocardial ischaemic episodes should take into account its non-stationary behaviour. The wavelet transform is an alternative tool for the analysis of non-stationary signals. Fourteen patients have been analysed, ranging from 40 to 64 years old and selected from the European Electrocardiographic ST-T Database (ESDB). These records contain 33 ST episodes, according to the notation of the ESDB, with durations of between 40s and 12 min. A method for analysing HRV signals using the wavelet transform was applied to obtain a time-scale representation for very low-frequency (VLF), low-frequency (LF) and high-frequency (HF) bands using the orthogonal multiresolution pyramidal algorithm. The design and implementation using fast algorithms included a specially adapted decomposition quadrature mirror filter bank for the frequency bands of interest. Comparing a normality zone against the ischaemic episode in the same record, increases in LF (0.0112 +/- 0.0101 against 0.0175 +/- 0.0208 s2 Hz(-1); p<0.1) and HF (0.0011 +/- 0.0008 against 0.00 17 +/- 0.0020 s2 Hz(-1); p<0.05) were obtained. The possibility of using these indexes to develop an ischaemic-episode classifier was also tested. Results suggest that wavelet analysis provides useful information for the assessment of dynamic changes and patterns of HRV during myocardial ischaemia.

Bionic wavelet transform: a new time-frequency method based on an auditory model

In this paper, a new adaptive wavelet transform, named bionic wavelet transform (BWT), is developed based on a model of the active auditory system. The most distinguishing characteristic of BWT is that its resolution in the time-frequency domain can be adaptively adjusted not only by the signal frequency but also by the signal instantaneous amplitude and its first-order differential. The automatically adjusted resolution, even in a fixed frequency along the time-axis, is achieved by introducing the active control of the auditory system into the wavelet transform (WT). Other properties of BWT include that: 1) BWT is a nonlinear transform that has high sensitivity and frequency selectivity; 2) BWT represents the signal with a concentrated energy distribution; and 3) the inverse BWT can reconstruct the original signal from its time-frequency representation. In order to compare these three properties between BWT and WT, experiments were conducted on both constructed signals and real speech signals. The results show that BWT performs better than WT in these three aspects, and that BWT is appropriate for speech signal processing, especially for cochlear implants.

Wavelet decomposition analysis of heart rate variability in aerobic athletes

Heart rate variability (HRV) can be quantified, among others, in the frequency domain using digital signal processing (DSP) techniques. The wavelet transform is an alternative tool for the analysis of non-stationary signals. The implementation of perfect reconstruction digital filter banks leads to multi resolution wavelet analysis. Software was developed in LabVIEW. In this study, the average power was compared at each decomposition level of a tachogram, containing the consecutive RR-intervals of two groups of subjects: aerobic athletes and a control group. Compared to the controls, the aerobic athletes showed an increased power in all frequency bands. These results are in accordance with values obtained by spectral analysis using the Fourier transform, suggesting that wavelet analysis could be an appropriate tool to evaluate oscillating components in HRV, but in addition to classic methods, it also gives a time resolution.

Automatic ectopic beat elimination in short-term heart rate variability measurement

Our studies deal with fully automatic measurement of heart rate variability (HRV) in short term electrocardiograms. Presently, all existing HRV analysis programs require user intervention for ectopic beat identification, especially of supraventricular ectopic beats (SVE). This makes the HRV measurement in large, e.g. epidemiological studies problematic. In this paper, we present a fully automatic algorithm for the discrimination of the ventricular (VE) and SVE ectopic beats from the normal QRS complexes suited for a reliable HRV analysis. The QRS identification is based on the template matching method. The ectopic beats are identified based on several morphological and timing properties of the electrocardiogram (ECG) signal. The method incorporates several approaches and makes HRV analysis of large numbers of electrocardiograms feasible. It uses the template matching for the basic morphology check of the QRS complex and the P-wave, the timing information to avoid unnecessary ectopic beat checks and to adjust thresholds and it also looks for a special QRS morphology, which is common in VEs. We used a testing set of 69 electrocardiograms selected from a large number of recordings. The selected ECGs contained abnormalities including ectopic beats, right branch bundle block, respiratory arrhythmia, blocked atrial extrasystole, high amplitude and wide T-waves. The evaluation of our method showed a specificity of 0.99, supraventricular ectopic beat sensitivity of 0.99 and ventricular ectopic beat sensitivity of 0.98.

A novel wavelet transform based analysis reveals hidden structure in ventricular fibrillation

We report a new method of interrogating the surface ECG signal using techniques developed in the field of wavelet transform analysis. Previously unreported structure within the ECG during ventricular fibrillation (VF) is found using a high-resolution decomposition of the signal employing the continuous wavelet transform. We believe that wavelet transform methods could lead to the development of powerful tools for use in the resuscitation of patients with cardiac arrest.

Wavelet transform to quantify heart rate variability and to assess its instantaneous changes

Heart rate variability is a recognized parameter for assessing autonomous nervous system activity. Fourier transform, the most commonly used method to analyze variability, does not offer an easy assessment of its dynamics because of limitations inherent in its stationary hypothesis. Conversely, wavelet transform allows analysis of nonstationary signals. We compared the respective yields of Fourier and wavelet transforms in analyzing heart rate variability during dynamic changes in autonomous nervous system balance induced by atropine and propranolol. Fourier and wavelet transforms were applied to sequences of heart rate intervals in six subjects receiving increasing doses of atropine and propranolol. At the lowest doses of atropine administered, heart rate variability increased, followed by a progressive decrease with higher doses. With the first dose of propranolol, there was a significant increase in heart rate variability, which progressively disappeared after the last dose. Wavelet transform gave significantly better quantitative analysis of heart rate variability than did Fourier transform during autonomous nervous system adaptations induced by both agents and provided novel temporally localized information.

Assessment of intersignal variability for discrimination of atrial fibrillation from atrial flutter

The analysis of endocardial signals obtained from an electrode located in the right atrium enabled by new dual chamber implantable cardioverter defibrillators may be helpful to provide additional therapies such as overdrive pacing or low energy atrial cardioversion for the treatment of concomitant atrial flutter (AFL) or atrial fibrillation (AF). Algorithms for discrimination of atrial tachyarrhythmias based on rate counting are of limited efficacy. The aim of this study was to assess the intersignal variability by using fast discrete wavelet transforms (FDWT) as a new method of discrimination of AF from AFL. Patients with spontaneous episodes of AF/AFL or patients who developed AF/AFL during an electrophysiological study were studied. The endocardial signals were recorded from the high right atrium using a transvenous 5 Fr bipolar electrode catheter (interelectrode spacing: 1 cm). The signals were digitized (2 kHz, 12-bit resolution) after amplification and filtering (40-500 Hz). Within data segments of 10-second duration, 25 consecutive signals were selected and normalized and FDWT was applied. Standard deviations of the wavelet coefficients (SD) from coarse scales (scale 4-8) were calculated. A total of 94 data segments (AF: 52, AFL: 42) from 28 patients were analyzed. SD at each considered scale was higher for AF than for AFL (P < 0.001). SD at scale 8 discriminated between AF from AFL with 100% sensitivity and specificity. We conclude that assessment of intersignal variability of bipolar endocardial recordings using FDWT is an effective method for the discrimination of AF from AFL. The implementation of this tool in a discrimination algorithm of an implantable device may help provide the appropriate differential therapy for atrial tachyarrhythmias.