Estimating Respiratory Pattern Variability by Symbolic Dynamics

Objectives: The traditional techniques of data analysis are often not sufficient to characterize the complex dynamics of respiration. In this study the respiratory pattern variability was analyzed using symbolic dynamics.

Methods: A group of 20 patients on weaning trials from mechanical ventilation were studied at two different pressure support ventilation levels. Breath duration (TTOT) time series and the relation TI/TTOT, that contains the influence of inspiratory time (TI), were considered. Length-3 words and 3 different symbols were proposed. The incidence of the overlapping τ and the parameter α were analyzed.

Results: From the breath duration time series, the distribution of words with probability of occurrence higher than 6% was concentrated on one word for low respiratory variability, whereas high variability was characterized by 4 words, presenting a statistically significant difference (p ≤ 0.0005). The probability occurrence of words “110” and “111” was also significantly different (p ≤ .0005) when comparing both variabilities.

Conclusion: The analysis carried out obtained discriminant functions able to correctly classify all the testing set series. These results permit the consideration of symbolic dynamics as a promising methodology to study the respiratory pattern variability.

Classification of patients undergoing weaning from mechanical ventilation using the coherence between heart rate variability and respiratory flow signal

Weaning from mechanical ventilation is still one of the most challenging problems in intensive care. Unnecessary delays in discontinuation and weaning trials that are undertaken too early are both undesirable. This study investigated the contribution of spectral signals of heart rate variability (HRV) and respiratory flow, and their coherence to classifying patients on weaning process from mechanical ventilation. A total of 121 candidates for weaning, undergoing spontaneous breathing tests, were analyzed: 73 were successfully weaned (GSucc), 33 failed to maintain spontaneous breathing so were reconnected (GFail), and 15 were extubated after the test but reintubated within 48 h (GRein). The power spectral density and magnitude squared coherence (MSC) of HRV and respiratory flow signals were estimated. Dimensionality reduction was performed using principal component analysis (PCA) and sequential floating feature selection. The patients were classified using a fuzzy K-nearest neighbour method. PCA of the MSC gave the best classification with the highest accuracy of 92% classifying GSucc versus GFail patients, and 86% classifying GSucc versus GRein patients. PCA of the respiratory flow signal gave the best classification between GFail and GRein patients (79% accuracy). These classifiers showed a good balance between sensitivity and specificity. Besides, the spectral coherence between HRV and the respiratory flow signal, in patients on weaning trial process, can contribute to the extubation decision.

Auditory and Nociceptive Stimuli Responses in the Electroencephalogram. A Non-linear Measures and Time-frequency Representation Based Analysis

INTRODUCTION

This article is part of the Focus Theme of Methods of Information in Medicine on "Biosignal Interpretation: Advanced Methods for Neural Signals and Images".

OBJECTIVES

An efficient way to investigate the neural basis of nociceptive responses is the analysis of the event-related brain potentials (ERPs). The main objective of this work was to study how adaptation and fatigue affect the ERPs to stimuli of different modalities, by characterizing the responses to infrequent and frequent stimulation in different recording periods.

METHODS

In this work, series of averaged EEG epochs recorded after thermal, electrical and auditory stimulation were analyzed with time-frequency representation and non-linear measures as spectral entropy and auto-mutual information function. The study was performed by considering the traditional EEG frequency bands.

RESULTS

The defined measures presented a statistical significance p-value < 0.01 and accuracy higher than 60% by differentiating windows of response to infrequent (I) and frequent (F) stimuli between the start and end of the EEG recording.

CONCLUSIONS

These measures permitted to observe some aspects of the subject's adaptation and the nociceptive response.

Complexity of the autonomic heart rate control in coronary artery occlusion in patients with and without prior myocardial infarction

Autonomic nervous system (ANS) is governed by complex interactions arising from feedback loops of nonlinear systems that operate over a wide range of temporal and spatial scales, enabling the organism to adapt to stress, metabolic changes and diseases. This study is aimed to assess multifractal and nonlinear characteristics of the ANS during ischemic events provoked by a prolonged percutaneous coronary intervention (PCI) procedure. Eighty-seven patients from the STAFF III database were used. Patients were classified into 2 groups: (1) with prior myocardial infarction (MI) and (2) without MI (noMI). R-R signals during three 3-min stages of the procedures were analyzed using multifractal and surrogate data techniques. Multifractal indices increased significantly from the pre-inflation stage to the post-deflation stage. These variations were more marked for the noMI group. Multifractal changes significantly correlated with both the decreased parasympathetic and the increased sympathetic modulations accounted by classical linear indices. Multifractal measures resulted to be a more powerful indicator than linear HRV indices in quantifying the ischemia-induced changes. Right coronary artery (RCA) occlusions provoke greater multifractal reactions throughout the PCI procedure. Our findings suggest reduced complex multifractal and nonlinear reactions of ANS activity in patients with prior MI in comparison to the noMI group, possibly due to degradation in the complexity of control mechanism of heart rate generation.

Symbolic dynamic analysis of relations between cardiac and breathing cycles in patients on weaning trials

Traditional time-domain techniques of data analysis are often not sufficient to characterize the complex dynamics of the cardiorespiratory interdependencies during the weaning trials. In this paper, the interactions between the heart rate (HR) and the breathing rate (BR) were studied using joint symbolic dynamic analysis. A total of 133 patients on weaning trials from mechanical ventilation were analyzed: 94 patients with successful weaning (group S) and 39 patients that failed to maintain spontaneous breathing (group F). The word distribution matrix enabled a coarse-grained quantitative assessment of short-term nonlinear analysis of the cardiorespiratory interactions. The histogram of the occurrence probability of the cardiorespiratory words presented a higher homogeneity in group F than in group S, measured with a higher number of forbidden words in group S as well as a higher number of words whose probability of occurrence is higher than a probability threshold in group S. The discriminant analysis revealed the best results when applying symbolic dynamic variables. Therefore, we hypothesize that joint symbolic dynamic analysis provides enhanced information about different interactions between HR and BR, when comparing patients with successful weaning and patients that failed to maintain spontaneous breathing in the weaning procedure.

SNP sets selection under mutual information criterion, application to F7/FVII dataset

One of the main goals of human genetics is to find genetic markers related to complex diseases. In blood coagulation process, it is known that genetic variability in F7 gene is the most responsible for observed variations in FVII levels in blood. In this work, we propose a method for selecting sets of Single Nucleotide Polymorphisms (SNPs) significantly correlated with a phenotype (FVII levels). This method employs a feature selection algorithm (variant of Sequential Forward Selection, SFS) based on a criterion of statistical significance of a mutual information functional. This algorithm is applied to a sample of independent individuals from the GAIT project. Main SNPs found by the algorithm are in correspondence with previous results published using family-based techniques.

DNA binding site characterization by means of Rényi entropy measures on nucleotide transitions

In this work, parametric information-theory measures for the characterization of binding sites in DNA are extended with the use of transitional probabilities on the sequence. We propose the use of parametric uncertainty measures such as Rényi entropies obtained from the transition probabilities for the study of the binding sites, in addition to nucleotide frequency-based Rényi measures. Results are reported in this work comparing transition frequencies (i.e., dinucleotides) and base frequencies for Shannon and parametric Rényi entropies for a number of binding sites found in E. Coli, lambda and T7 organisms. We observe that the information provided by both approaches is not redundant. Furthermore, under the presence of noise in the binding site matrix we observe overall improved robustness of nucleotide transition-based algorithms when compared with nucleotide frequency-based method.

Information flow to assess cardiorespiratory interactions in patients on weaning trials

Nonlinear processes of the autonomic nervous system (ANS) can produce breath-to-breath variability in the pattern of breathing. In order to provide assess to these nonlinear processes, nonlinear statistical dependencies between heart rate variability and respiratory pattern variability are analyzed. In this way, auto-mutual information and cross-mutual information concepts are applied. This information flow analysis is presented as a short-term non linear analysis method to investigate the information flow interactions in patients on weaning trials. 78 patients from mechanical ventilation were studied: Group A of 28 patients that failed to maintain spontaneous breathing and were reconnected; Group B of 50 patients with successful trials. The results show lower complexity with an increase of information flow in group A than in group B. Furthermore, a more (weakly) coupled nonlinear oscillator behavior is observed in the series of group A than in B.

Patients on weaning trials from mechanical ventilation classified with neural networks and feature selection

One of the challenges in intensive care is the process of weaning from mechanical ventilation. We studied the differences in respiratory pattern variability between patients capable of maintaining spontaneous breathing during weaning trials and patients that fail to maintain spontaneous breathing. In this work, neural networks were applied to study these differences. 64 patients from mechanical ventilation are studied: Group S with 32 patients with Successful trials and Group F with 32 patients that Failed to maintain spontaneous breathing and were reconnected. A performance of 64.56% of well classified patients was obtained using a neural network trained with the whole set of 35 features. After the application of a feature selection procedure (backward selection) 84.56% was obtained using only 8 of the 35 features.

Heart rate variability analysis based on time-frequency representation and entropies in hypertrophic cardiomyopathy patients

In hypertrophic cardiomyopathy (HCM) patients there is an increased risk of premature death, which can occur with little or no warning. Furthermore, classification for sudden cardiac death on patients with HCM is very difficult. The aim of our study was to improve the prognostic value of heart rate variability (HRV) in HCM patients, giving insight into changes of the autonomic nervous system. In this way, the suitability of linear and nonlinear measures was studied to assess the HRV. These measures were based on time-frequency representation (TFR) and on Shannon and Rényi entropies, and compared with traditional HRV measures. Holter recordings of 64 patients with HCM and 55 healthy subjects were analyzed. The HCM patients consisted of two groups: 13 high risk patients, after aborted sudden cardiac death (SCD); 51 low risk patients, without SCD. Five-hour RR signals, corresponding to the sleep period of the subjects, were considered for the analysis as a comparable standard situation. These RR signals were filtered in the three frequency bands: very low frequency band (VLF, 0-0.04 Hz), low frequency band (LF, 0.04-0.15 Hz) and high frequency band (HF, 0.15-0.45 Hz). TFR variables based on instantaneous frequency and energy functions were able to classify HCM patients and healthy subjects (control group). Results revealed that measures obtained from TFR analysis of the HRV better classified the groups of subjects than traditional HRV parameters. However, results showed that nonlinear measures improved group classification. It was observed that entropies calculated in the HF band showed the highest statistically significant levels comparing the HCM group and the control group, p-value < 0.0005. The values of entropy measures calculated in the HCM group presented lower values, indicating a decreasing of complexity, than those calculated from the control group. Moreover, similar behavior was observed comparing high and low risk of premature death, the values of the entropy being lower in high risk patients, p-value < 0.05, indicating an increase of predictability. Furthermore, measures from information entropy, but not from TFR, seem to be useful for enhanced risk stratification in HCM patients with an increased risk of sudden cardiac death.

Association between short term and long term communication in pathological autonomic control

Autonomic Information Flow (AIF) reflects the time scale dependence of autonomic communications such as vagal, sympathetic, and slower rhythms and their complex interplay. We investigated the hypothesis that pathologically disturbed short term control is associated with simplified complex long term control. This particular characteristic of altered autonomic communication was evaluated in different medical patient groups. Holter recordings were assessed in patients with multiple organ dysfunction (MODS) (26 survivors, 10 non-survivors); with heart failure (14 low risk-without history of aborted cardiac arrest (CA), 13 high risk--with history of CA); with idiopathic dilated cardiomyopathy (IDC) (26 low risk, 11 high risk of CA), after myocardial infarction (MI) (1221 low risk--survivors, 55 high risk--non-survivors); after abdominal aorta surgery (AAS, 32 length of stay in hospital LOS>7 days, 62 LOS < or =7 days). AIF of short and long time scales was investigated. We found a fundamental association of increased short term randomness and decreased long term randomness due to pathology. Concerning risk, high risk patients were characterized by increased short term complexity and decreased long term complexity in all patients groups with the exception of the IDC patients. We conclude that different time scales of AIF represent specific pathophysiological aspects of altered autonomic communication and control. The association of altered short term control with simplified long term behavior might be a pathophysiologically relevant compensation mechanism in the case of a disturbed fastest actuator. This knowledge might be useful for the development of comprehensive therapeutic strategies besides the predictive implications.

Study of the respiratory pattern variability in patients during weaning trials

Mechanical ventilators are used to provide life support in patients with respiratory failure. One of the challenges in intensive care is the process of weaning from mechanical ventilation. We studied the differences in respiratory pattern variability between patients capable of maintaining spontaneous breathing during weaning trials and patients that fail to maintain spontaneous breathing. The respiratory pattern was characterized by the following time series: inspiratory time (T(I)), expiratory time (T(E)), breath duration (T(Tot)), tidal volume (V(T)), fractional inspiratory time (T(I)/T(Tot)), mean inspiratory flow (V(T)/T(I)), respiratory frequency (f), and rapid shallow breathing index (f/V(T)). The variational activity of breathing was partitioned into autoregressive, periodic and white noise fractions. Patients with unsuccessful trial presented a tendency to higher values of gross variability of V(T)/T(I) and f/V(T), and lower values of T(I). The autocorrelation coefficients tended to present higher values for T(I), T(I)/T(Tot) and V(T)/T(I). During both successful and unsuccessful T-tube test uncorrelated random behavior constituted > 75% of the variance of each time breath components and represented 50 to 70% in the breath component related to V(T). Correlated behavior represented 6 to 21% in time components and 28 to 50% in component related to V(T).

Hidden markov models and mutual information analysis to characterize nonlinear dynamics in heart rate variability

A study of nonlinear dynamics of the heart rate variability (HRV) was performed using hidden Markov models (HMM) and Mutual Information (MI). A methodology based on HMM has been developed in the present work. Cardiac RR series were analyzed in the three frequency bands: HF (0.15-0.45Hz), high frequency band; LF (0.04-0.15Hz), low frequency band; VLF (0.003-0.04Hz), very low frequency band. These series (0, observations) were modeled using HMM. The model &#955;=(A,B,&#8719;) was selected so that P(O/&#955;) was locally maximized. Ergodic topology and N=10 states were also considered for this analysis. Different measures based on HMM were defined and obtained from RR time series of 37 Idiopathic Dilated Crdiomyopathy (IDC) patients and 46 healthy subjects (NRM), during awake and sleep stages. Two groups of IDC patients were considered: 11 high risk (HR) patients, after aborted sudden cardiac death (SCD) or who died during the follow up; 26 low risk (LR) patients, without SCD. Some HMM measures showed high percentages (up to 100%) of well classified subjects in all groups.

Joint symbolic dynamic analysis of cardiorespiratory interactions in patients on weaning trials

Assessing autonomic control provides information about patho-physiological imbalances. Measures of variability of the cardiac interbeat duration RR(n) and the variability of the breath duration T<inf>Tot</inf>(n) are sensitive to those changes. The interactions between RR(n) and T<inf>Tot</inf>(n) are complex and strongly non-linear. A study of joint symbolic dynamics is presented as a new short-term non-linear analysis method to investigate these interactions in patients on weaning trials. 78 patients from mechanical ventilation are studied: Group A (patients that failed to maintain spontaneous breathing and were reconnected) and Group B (patients with successful trials). Using the concept of joint symbolic dynamics, cardiac and respiratory changes were transformed into a word series, and the probability of occurrence of each word type was calculated and compared between both groups. Significant differences were found in 13 words, and the most significant p<inf>n</inf>(W<inf>c010, r010</inf>): 0.0041 &#177; 0.0036 (group A) against 0.0012 &#177; 0.0024 (group B), p-value = 0.00001. The number of seldom occurring word types (forbidden words) also presents significant differences fw<inf>cr</inf>: 6.9 &#177; 6.6 against 13.5 &#177; 5.3, p-value = 0.00004. Joint symbolic dynamics provides an efficient non-linear representation of cardiorespiratory interactions that offers simple physiological interpretations.

Evaluation of different rhythms by hidden Markov models in heart rate variability of hypertrophic cardiomyopathy patients

Synchronization and regularity between different rhythms were evaluated in the HRV using hidden Markov models (HMMs) at very low (VLF), low (LF) and high (HF) frequency bands. Phase synchronization of these rhythms was studied in RR series of hypertrophic cardiomyopathy (HCM) patients during the sleeping period. Two groups of patients were considered in the HCM group: high risk (HR), patients after aborted sudden death (SCD) or that died during follow up, and low risk (LR), patients without SCD. RR time-series were filtered in the following frequency-bands, VLF, LF and HF. The RR phase differences of HF vs. VLF, HF vs. LF and LF vs. VLF were calculated and then the amplitude range partitioned into 8 bins. Finally, these series (O, observations) were modeled using HMM. The models lambda = (A,B,pi) were selected such that P(O/lambda) was locally maximized. Ergodic topology and N = {5,10,15,20} states were considered also for this analysis. Ergodic HMMs with 10 states were found to be sufficient to characterize the HRV rhythms of HR and LR patients. Different synchronization strength was observed studying the phase entropies. However, only the parameters obtained from the HMM were able to differentiate the different groups, with p-value < 0.0005.

Modeling and evaluation of respiratory and muscle pattern during hypercapnic stimulus

Understanding the respiratory control system and the ventilatory pattern under hypercapnic stimulus is important to interpret the acute exacerbation of COPD and the condition of patients connected to mechanical ventilation. The purpose of this study is the analysis of respiratory and muscle parameters in order to obtain the most sensitive and characteristic of different levels of hypercapnic stimulus. Parameters defined and calculated from pressure signals show the highest variations with the increment of stimulus. Other ones like exhaled ventilation or ratios between respiratory parameters are more influenced by hypercapnia than tidal volume, respiratory frequency or even end tidal CO2. Muscle parameters from electromyographic signals of three respiratory muscles are calculated in time and frequency domain. In spite of greater variability between subjects, the most interesting muscles because of their activation with higher stimulus are in the following order: diaphragm, sternomastoid and genioglossus. Moreover, a model of respiratory control system is evaluated in order to predict and simulate appropriately this ventilatory stimulus. In spite of scattered real data, they are compared with simulation results obtained by the model and predicted by means of a specific respiratory optimization.

Analysis of the nonlinear autodependencies of respiratory pattern in patients on weaning trials

Traditional time domain techniques of data analysis are often not sufficient to characterize the nonlinear dynamics of respiration. In this study, the respiratory pattern variability was analyzed using auto mutual information measures. These provide access to nonlinear statistical autodependencies of respiratory pattern variability. A group of 20 patients on weaning trials from mechanical ventilation were studied at two different pressure support ventilation levels, in order to obtain respiratory volume signals with different variability. Time series of breathing duration, inspiratory time, fractional inspiratory time, tidal volume and mean inspiratory flow were analyzed. Different measures based on auto-mutual information were studied to characterize the respiratory pattern variability with regard to its complex organization.

Hidden Markov Models Based on Symbolic Dynamics for Statistical Modeling of Cardiovascular Control in Hypertensive Pregnancy Disorders

Discrete hidden Markov models (HMMs) were applied to classify pregnancy disorders. The observation sequence was generated by transforming RR and systolic blood pressure time series using symbolic dynamics. Time series were recorded from 15 women with pregnancy-induced hypertension, 34 with preeclampsia and 41 controls beyond 30th gestational week. HMMs with five to ten hidden states were found to be sufficient to characterize different blood pressure variability, whereas significant classification in RR-based HMMs was found using fifteen hidden states. Pregnancy disorders preeclampsia and pregnancy induced hypertension revealed different patho-physiological autonomous regulation supposing different etiology of both disorders.

Optimized Symbolic Dynamics Approach for the Analysis of the Respiratory Pattern

Traditional time domain techniques of data analysis are often not sufficient to characterize the complex dynamics of respiration. In this paper, the respiratory pattern variability is analyzed using symbolic dynamics. A group of 20 patients on weaning trials from mechanical ventilation are studied at two different pressure support ventilation levels, in order to obtain respiratory volume signals with different variability. Time series of inspiratory time, expiratory time, breathing duration, fractional inspiratory time, tidal volume and mean inspiratory flow are analyzed. Two different symbol alphabets, with three and four symbols, are considered to characterize the respiratory pattern variability. Assessment of the method is made using the 40 respiratory volume signals classified using clinical criteria into two classes: low variability (LV) or high variability (HV). A discriminant analysis using single indexes from symbolic dynamics has been able to classify the respiratory volume signals with an out-of-sample accuracy of 100%.

Variability analysis of the respiratory volume based on non-linear prediction methods

This work proposed and studied a method of automatically classifying respiratory volume signals as high or low variability by means of non-linear analysis of the respiratory volume. The analysis used volume signals generated by the respiratory system to construct a model of its dynamics and to estimate the quality of the predictions made with the model. Different methods of prediction evaluation, prediction horizons and embedding dimensions were also analysed. Assessment of the method was made using a database that contained 40 respiratory volume signals classified using clinical criteria into two classes: low or high variability. The results obtained using the method of surrogate data provided evidence of non-linear determinism in the respiratory volume signals. A discriminant analysis carried out using non-linear prediction variables classified the respiratory volume signals with an accuracy of 95%.

Estimating respiratory pattern variability by symbolic dynamics

OBJECTIVES

The traditional techniques of data analysis are often not sufficient to characterize the complex dynamics of respiration. In this study the respiratory pattern variability was analyzed using symbolic dynamics.

METHODS

A group of 20 patients on weaning trials from mechanical ventilation were studied at two different pressure support ventilation levels. Breath duration (T(TOT)) time series and the relation T(I)/T(TOT), that contains the influence of inspiratory time (T(I)), were considered. Length-3 words and 3 different symbols were proposed. The incidence of the overlapping tau and the parameter alpha were analyzed.

RESULTS

From the breath duration time series, the distribution of words with probability of occurrence higher than 6% was concentrated on one word for low respiratory variability, whereas high variability was characterized by 4 words, presenting a statistically significant difference (p </= 0.0005). The probability occurrence of words "110" and "111" was also significantly different (p</= 0.0005) when comparing both variabilities.

CONCLUSION

The analysis carried out obtained discriminant functions able to correctly classify all the testing set series. These results permit the consideration of symbolic dynamics as a promising methodology to study the respiratory pattern variability.

Comparison of auditory evoked potentials and the A-line ARX Index for monitoring the hypnotic level during sevoflurane and propofol induction

BACKGROUND

Extraction of the middle latency auditory evoked potentials (AEP) by an auto regressive model with exogenous input (ARX) enables extraction of the AEP within 1.7 s. In this way, the depth of hypnosis can be monitored at almost real-time. However, the identification and the interpretation of the appropriate signals of the AEP could be difficult to perform during the anesthesia procedure. This problem was addressed by defining an index which reflected the peak amplitudes and latencies of the AEP, developed to improve the clinical interpretation of the AEP. This index was defined as the A-line Arx Index (AAI).

METHODS

The AEP and AAI were compared with the Modified Observers Assessment of Alertness and Sedation Scale (MOAAS) in 24 patients scheduled for cardiac surgery, anesthetized with propofol or sevoflurane.

RESULTS

When comparing the AEP peak latencies and amplitudes and the AAI, measured at MOAAS level 5 and level 1, significant differences were achieved. (mean(SD) Nb latency: MOAAS 5 51.1 (7.3) ms vs. MOAAS 1: 68.6 (8.1) ms; AAI: MOAAS 5 74.9 (13.3) vs. MOAAS 1 20.7 (4.7)). Among the recorded parameters, the AAI was the best predictor of the awake/anesthetized states.

CONCLUSION

We conclude that both the AAI values and the AEP peak latencies and amplitudes correlated well with the MOAAS levels 5 (awake) and 1 (anesthetized).

Influence of estimators of spectral density on the analysis of electromyographic and vibromyographic signals

Electromyographic (EMG) and vibromyographic (VMG) signals are related to electrical and mechanical muscle activity, respectively. It is known that variations in their frequency components are related to changes in muscle activity and fatigue. The aims of this study were: (1) to analyse the resolution, variance and bias of different estimations of power spectral density function (PSD); and (2) to evaluate the influence of the spectral estimation method on three indices calculated from the PSD of EMG and VMG signals: mean (f(m)) and median (f(c)) frequencies and the ratio of high and low frequency components (H/L ratio) to select the most suitable estimator. Myographic signals were recorded from the sternomastoid muscle, an accessory respiratory muscle, during breathing. For non-parametric methods, Welch periodograms and correlograms were analysed with different windows. Autoregressive (AR) moving average (MA) and ARMA models with different orders were evaluated in the parametric methods. The reproducibility of the results was also studied. Frequency indices, particularly the H/L ratio and f(c), changed considerably when varying the following parameters of the estimators: periodogram with segment durations longer than 150 ms in EMG and with any duration in VMG signals; correlogram with window length shorter than 10% of the total number of samples; and AR models with an order lower than 10, 20 and 40 in f(c), fm and H/L ratio, respectively, in both myographic signals.

[Comparison of an auditory evoked potentials index and a bispectral index versus clinical signs for determining the depth of anesthesia produced by propofol or sevoflurane]

OBJECTIVES

To evaluate an anesthetic depth index (ADI) obtained from auditory evoked potentials and a bispectral EEG index (BIS) in comparison with clinical assessment of anesthetic depth using the modified observer's assessment of awareness/sedation scale (MOAA/SS), for induction of anesthesia with propofol or sevoflurane as the only agent.

PATIENTS AND METHODS

The ADI and BIS were recorded simultaneously in this prospective study and compared to the MOAA/SS during the anesthetic induction of 26 adults undergoing elective heart surgery. Assignment of patients to two groups was random. Group A (n = 13) patients were induced with propofol (target dose 5 micrograms.ml-1 in 5 min). Induction in group B (n = 13) was with sevoflurane (8% tidal volume). A scheme of awake-sleeping-awake-sleeping was followed. The means of the two indexes were compared (Mann-Whitney test) one minute before the patient slept (awake) and one minute later (sleeping), and the evolution of the indexes was compared during awake/sleep and sleep/awake phase changes and while the patients were in a stable sleep phase. The sensitivity and specificity of each index was analyzed in function of the MOAA/SS. We also analyzed the time elapsing from the moment the patient fell asleep (MOAA/SS 2) until the two indexes reached published reference values (ADI = 38, BIS = 60).

RESULTS

After induction with propofol (group A) the ADI fell to 29.2 +/- 11.7 and the BIS fell to 63.5 +/- 13.4. After induction with sevoflurane (group B) the ADI fell to 33.8 +/- 14.9 and the BIS to 66.8 +/- 15. The ADI value that best discriminated between arousal and sleeping (sensitivity 100%) was 38; the BIS value that best discriminated was 60. The responses to sound in decibels (dB) during "awake/sleeping" and "sleeping/awake" phases were, respectively, -3.8 dB and -4.5 dB for the ADI and -1.5 dB and -0.8 dB for the BIS. With the patient in stable sleep, response to the two indexes was at -0.79 dB. In group A, the ADI detected MOAA/SS 2 significantly earlier (ADI 13.1 +/- 30 s; BIS 56 +/- 36 s; p < 0.05). No patient reported remembering the study period.

CONCLUSIONS

Monitoring anesthetic depth with the ADI or BIS was technically easy and effective for detecting whether patients were awake or sleeping. The ADI response was faster and identified awake/sleeping and sleeping/awake phase changes better than did the BIS.

Time-frequency analysis of the RT and RR variability to stratify hypertrophic cardiomyopathy patients

The RT interval is a measure of the ventricular repolarization and is partially influenced by the sympathovagal balance. The analysis of the variation of the duration of the RT and RR intervals might bring new information about the arrhythmogenic vulnerability and autonomic imbalance. The RR signal and its spectral density (SD) are characterized by two different patterns during the sleep period. On the basis of this information, RT and RR sequences have been automatically classified into two patterns, R and N. In this work, we propose a methodology to define new variables that are able to distinguish patients with hypertrophic cardiomyopathy (HCM) who later developed sudden cardiac death (SCD) from HCM patients without such episode during the follow-up. These variables are based on the instantaneous frequency calculation using time-frequency representation of the RT and RR signals previously classified into R and N patterns. In this study, three spectral bands have been considered: low-frequency band (LF, 0-0.07 Hz), mid-frequency band (MF, 0.07-0.15 Hz), and high-frequency band (HF, 0.15-0.45 Hz). Then a suitable combination of mean energy and mean frequency of the RT and RR signals in the MF and HF bands has allowed HCM patients with SCD to be discriminated from HCM patients without SCD (P < 0.001).

Study of myographic signals from sternomastoid muscle in patients with chronic obstructive pulmonary disease

Analysis of the respiratory muscle activity is a promising technique for diagnosis of respiratory diseases, such as chronic obstructive pulmonary disease (COPD). The sternomastoid muscle (SMM) was selected to study the activity of respiratory muscles due to its accessibility in order to define a noninvasive analysis. The aims of this work are two: analyze the relationship between the SMM function and pulmonary obstruction, and study the influence of spectral estimator on frequency parameters related with the muscle activity. For the first goal, we propose the analysis of vibromyographic and electromyographic signals from the SMM to study the muscle function during two ventilatory tests. Activity of SMM was found by means of several indexes: root-mean-square (rms) values, mean and median frequencies, and ratio between high and low-frequency components. For the second goal, spectral analysis was performed by means of nonparametric methods: Correlogram and Welch periodogram, and parametric methods: autoregressive (AR), moving average (MA), and ARMA models. It is deduced that these indexes show muscle activity and certain fatigue of the SMM, whose muscle function depends on the level of pulmonary obstruction, and they depend a lot of spectral estimator being the more suitable an AR model with high order.

Analysis of tracheal sounds during forced exhalation in asthma patients and normal subjects: bronchodilator response effect

PURPOSE

During the past 10 years, the acoustic analysis of breath sounds has been used as a diagnostic tool in patients suffering from obstructive respiratory diseases. Acoustic analysis might be able to monitor the response to bronchodilator therapy in a clinical setting. So far, few studies have been carried out in asthmatic patients. To assess the responses of a sampling of asthma patients to an inhaled bronchodilator (terbutaline) by means of spectral analysis of the tracheal sound performed during forced expiratory maneuvers.

MATERIAL AND METHODS

Seventeen nonsmoking asthma patients (9 were male, 8 were female) who had been suffering from the disease for > or = 15 years were included in the study, as were 15 normal subjects (7 were male, 8 were female). The average age (+/- SD) was 56.5 +/- 15.2 years (FVC, 2.7 +/- 0.9 L [63.4%]; FEV1, 1.5 +/- 0.6 L [53.0%]). The tracheal sounds were collected during three forced expiratory maneuvers with a sampling frequency of 5,000 Hz and were analyzed by applying a 16-parameter autoregressive model.

RESULTS

The centroid frequency decreased after the bronchodilator was given at different flow segments between 1.2 and 0.4 L/s, with significant changes between 0.6 and 0.4 L/s.

CONCLUSIONS

Patients with asthma showed changes in the spectral acoustic analysis frequencies after the administration of a bronchodilator drug (terbutaline) during forced expiratory maneuvers.

Identification of causal relations between haemodynamic variables, auditory evoked potentials and isoflurane by means of fuzzy logic

The aim of this study was to identify a possible relationship between haemodynamic variables, auditory evoked potentials (AEP) and inspired fraction of isoflurane (ISOFl). Two different models (isoflurane and mean arterial pressure) were identified using the fuzzy inductive reasoning (FIR) methodology. A fuzzy model is able to identify non-linear and linear components of a causal relationship by means of optimization of information content of available data. Nine young female patients undergoing hysterectomy under general anaesthesia were included. Mean arterial pressure (MAP), heart rate (HR), end-tidal expired carbon dioxide (CO2ET), AEP and ISOFl were monitored with a sampling time of 10 s. The AEP was extracted using an autoregressive model with exogenous input (ARX model) which decreased the processing time compared with a moving time average. The AEP was mapped into a scalar, termed the depth of anaesthesia index (DAI) normalized to 100 when the patient was awake and descending to an average of 25 during loss of consciousness. The FIR methodology identified those variables among the input variables (MAP, HR, CO2ET, DAI or ISOFl) that had the highest causal relation with the output variables (ISOFl and MAP). The variables with highest causal relation constitute the ISOFl and MAP models. The isoflurane model predicted the given anaesthetic dose with a mean error of 12.1 (SD 10.0)% and the mean arterial pressure model predicted MAP with a mean error of 8.5 (7.8)%.

Spectral analysis of heart period variance (HPV)--a tool to stratify risk following myocardial infarction

The purpose of this study was to contribute to the improvement of stratification of post-myocardial infarction patients at increased risk of malignant ventricular arrhythmia (MVA). Power spectral analysis of heart period variability (HPV) was used as a non-invasive tool to assess cardiac autonomic control. Three groups were used: (1) post-myocardial infarction patients with MVA; (2) post-myocardial infarction patients without MVA; and (3) a control group without heart disease. Spectral analysis of HPV (AR model) was performed on four minute long RR-interval time series derived from consecutive hours of Holter ECG. Significant decrease of powers of mid-frequency (MF) (70-150 mHz) and high-frequency (HF) (150-450 mHz) spectral components of HPV was obtained in Group 1 as compared to Group 2 (p = 0.001 and p = 0.02, respectively). There were no significant differences between groups concerning the power of low frequency (LF) (10-70 mHz) component HPV, spectra of patients in Group 1 were dominated by a single low frequency spectral peak (with a central frequency of 37 mHz). The relative power was computed as the percentage of power in each of the above (HF, MF, LF) components related to the total spectral power. Highly significant differences (p = 0.04) were obtained between Group 1 and Group 2 concerning relative powers of MF and LF components as well as LF/MF ratio. The above method appeared to be highly sensitive in differentiating patients with increased risk of MVA.

Analysis of abnormal signals within the QRS complex of the high-resolution electrocardiogram

This paper presents a new, quantitative approach to measuring abnormal intra-QRS signals, using the high-resolution electrocardiogram (HRECG). These signals are conventionally known as QRS "notches and slurs." They are measured qualitatively and form the basis for the ECG identification of myocardial infarction. The HRECG is used for detection of ventricular late potentials (LP), which are linked with the presence of a reentry substrate for ventricular tachycardia (VT) after a myocardial infarction. LP's are defined as signals from areas of delayed conduction which outlast the normal QRS period. Our objective is to quantify very low-level abnormal signals that may not outlast the normal QRS period. In this work, abnormal intra-QRS potentials (AIQP) were characterized by removing the predictable, smooth part of the QRS from the original waveform. This was represented as the impulse response of an ARX parametric model, with model order selected empirically from a training data set. AIQP were estimated using the residual of the modeling procedure. Critical AIQP parameters to separate VT and non-VT subjects were obtained using discriminant functions. Results suggest that AIQP indexes are a new predictive index of the HRECG for VT. The concept of abnormal intra-QRS potentials permits the characterization of pathophysiological signals contained wholly within the normal QRS period, but related to arrhythmogenesis. The new method may have other applications, such as detection of myocardial ischemia and improved ECG identification of the site of myocardial infarction, particularly in the absence of Q waves.

Automatic measurement of corrected QT interval in Holter recordings: comparison of its dynamic behavior in patients after myocardial infarction with and without life-threatening arrhythmias

This study was designed to determine the value of automatic corrected QT-interval measurement in Holter tapes in patients after myocardial infarction as a marker of life-threatening ventricular arrhythmias. We compared the corrected QT interval, automatically measured in 24-hour Holter recordings, in two groups of patients after myocardial infarction: group I was composed of 14 patients admitted consecutively to our hospital for documented sustained ventricular tachycardia or out-of-hospital cardiac arrest. Group II consisted of 28 patients with previous myocardial infarction with characteristics similar to those of group I, but without malignant ventricular arrhythmias in the follow-up. The global mean 24-hour corrected QT interval was longer in group I (425 +/- 20 msec) than in those patients after myocardial infarction without arrhythmias (group II) (405 +/- 17 msec; p < 0.01). Furthermore, a significant proportion of patients of group I (seven of 14) exhibited more peaks of corrected QT longer than 500 msec compared with patients of group II (two of 28; p < 0.005). A circadian rhythm of corrected QT peaks was observed in group I, having a significantly higher incidence from 11 PM to 11 AM (p < 0.05). We conclude that automatic corrected QT-interval measurement on Holter electrocardiogram is now available and feasible. Our results suggest that this is a marker for risk assessment of life-threatening ventricular arrhythmias. Large-scale trials are needed to confirm these results and to determine the predictive value of this technique for risk stratification.

Analysis of abnormal intra-QRS potentials. Improved predictive value for arrhythmic events with the signal-averaged electrocardiogram

BACKGROUND

Using the signal-averaged ECG (SAECG), this study developed a new electrical index for predicting arrhythmic events: abnormal intra-QRS potentials (AIQP).

METHODS AND RESULTS

We studied 173 patients followed after myocardial infarction for a mean duration of 14 +/- 7 months. Sixteen arrhythmic events occurred, defined as sudden cardiac death, documented sustained ventricular tachycardia, or non-fatal cardiac arrest. Noninvasive indices of arrhythmia risk were measured, including AIQP, conventional SAECG, Holter, and left ventricular ejection fraction (LVEF). Abnormal intra-QRS potentials were defined as abnormal signals occurring anywhere within the QRS period. They were estimated with a lead-specific, parametric modeling method that removed the smooth, predictable part of the QRS. AIQPs are characterized by the remaining transient, unpredictable component of the QRS and manifest as low-amplitude notches and slurs. A combined XYZ-lead AIQP index exhibited higher specificity (95%) and predictive value (PV) (+PV, 47%; -PV, 94%) than the conventional SAECG in combination with Holter and LVEF (specificity, 89%; +PV, 25%; -PV, 93%).

CONCLUSIONS

AIQP improved specificity and predictive value, compared with conventional tests, for prediction of arrhythmic events. AIQP emerged as the best noninvasive univariate predictor of arrhythmic events after myocardial infarction in this study. A review of several other reports shows that AIQP in the present study outperformed the conventional predictive indices reported in those other data sets.

Acoustic analysis of snoring sound in patients with simple snoring and obstructive sleep apnoea

Snoring, a symptom which may indicate the presence of the obstructive sleep apnoea syndrome (OSA), is also common in the general population. Recent studies have suggested that the acoustic characteristics of snoring sound may differ between simple snorers and OSA patients. We have studied a small number of patients with simple snoring and OSA, analysing the acoustic characteristics of the snoring sound. Seventeen male patients, 10 with OSA (apnoea/hypopnoea index (AHI) 26.2 events x h(-1)) and seven simple snorers (AHI 3.8 events x h(-1)), were studied. Full night polysomnography was performed and the snoring sound power spectrum was analysed. Spectral analysis of snoring sound showed the existence of two different patterns. The first pattern was characterized by the presence of a fundamental frequency and several harmonics. The second pattern was characterized by a low frequency peak with the sound energy scattered on a narrower band of frequencies, but without clearly identified harmonics. The seven simple snorers and two of the 10 patients with OSA (AIH 13 and 14 events x h(-1), respectively) showed the first pattern. The rest of the OSA patients showed the second pattern. The peak frequency of snoring was significantly lower in OSA patients, with all but one OSA patient and only one simple snorer showing a peak frequency below 150 Hz. A significant negative correlation was found between AHI and peak and mean frequencies of the snoring power spectrum (p<0.0016 and p<0.0089, respectively). In conclusion, this study demonstrates significant differences in the sound power spectrum of snoring sound between subjects with simple snoring and obstructive sleep apnoea patients.

Adaptive estimation of QRS complex wave features of ECG signal by the Hermite model

The most characteristic wave set in ECG signals is the QRS complex. Automatic procedures to classify the QRS are very useful in the diagnosis of cardiac dysfunctions. Early detection and classification of QRS changes are important in real-time monitoring. ECG data compression is also important for storage and data transmission. An Adaptive Hermite Model Estimation System (AHMES) is presented for on-line beat-to-beat estimation of the features that describe the QRS complex with the Hermite model. The AHMES is based on the multiple-input adaptive linear combiner, using as inputs the succession of the QRS complexes and the Hermite functions, where a procedure has been incorporated to adaptively estimate a width related parameter b. The system allows an efficient real-time parameter extraction for classification and data compression. The performance of the AHMES is compared with that of direct feature estimation, studying the improvement in signal-to-noise ratio. In addition, the effect of misalignment at the QRS mark is shown to become a neglecting low-pass effect. The results allow the conditions in which the AHMES improves the direct estimate to be established. The application is shown, for subsequent classification, of the AHMES in extracting the QRS features of an ECG signal with the bigeminy phenomena. Another application is highlighted that helps wide ectopic beats detection using the width parameter b.

Multiresolution wavelet analysis of the body surface ECG before and after angioplasty

Electrocardiographic recordings of patients with coronary artery stenosis, made before and after angioplasty, were analyzed by the multiresolution wavelet transform (MRWT) technique. The MRWT decomposes the signal of interest into its coarse and detail components at successively finer scales. MRWT was carried out on different leads in order to compare the P-QRS-T complex from recordings made before with those made after percutaneous transluminal coronary angioplasty (PTCA). ECG signals before and after successful PTCA procedures show distinctive changes at certain scales, thus helping to identify whether the procedure has been successful. In six patients who underwent right coronary artery PTCA, varying levels of reperfusion were achieved, and the changes in the detail components of ECG were shown to correlate with the successful reperfusion. The detail components at scales 5 and 6, corresponding approximately to the frequencies in the range of 2.3-8.3 Hz, are shown to be the most sensitive to ischemia-reperfusion changes (p < 0.05). The same conclusion was reached by synthesizing the post-PTCA signals from pre-PTCA signals with the help of these detail components. For on-line monitoring a vector plot, analogous to vector cardiogram, of the two most sensitive MRWT detail components is proposed. Thus, multiresolution analysis of ECG may be useful as a monitoring and diagnostic tool during angioplasty procedures.

Karhunen-Loève transform as a tool to analyze the ST-segment. Comparison with QT interval

The spatial and temporal courses of ventricular repolarization are quite sensitive to the biochemical and biophysiologic environment of the myocardial cells, and are therefore often an early marker of heart disease, particularly of ischemia. The detailed morphology of the surface electrocardiogram contains considerable information about the repolarization process. The ST-segment changes with ischemia, injury, and drugs. The QT interval is affected by drugs, heart rate, and autonomic tone, and in some situations may identify individuals at high risk for arrhythmias and sudden death. Variability in the shape, including duration, of the ST-T waves reflects autonomic nervous system activity and may identify high-risk patients. Automated methods for quantitatively characterizing ST-T complexes are important in studying long-term electrocardiographic records. Two computer-based measurement procedures for characterizing the repolarization period were comparatively analyzed: Karhunen-Loève (KL) transform representation of the ST-T shape and measurement of beat-to beat durations of repolarization (QT intervals). The results of KL transform representation and time-domain QT measurement algorithms for studying the repolarization period of the electrocardiogram on the European ST-T database are presented. It was found that about 20% of the records present a quasiperiodic KL pattern of ischemic ST-T activity and another 20% exhibit repetitive but not clearly periodic patterns of ischemic ST-T changes. From these ischemic records, 50% showed QT variations in at least one lead associated with the ischemic episodes.

Ventricular late potentials characterization in time-frequency domain by means of a wavelet transform

The main transforms of Cohen's class allow signal representation simultaneously in time and frequency domains. Wavelet transforms make it possible to link the temporal window width to the analyzing frequency and leads to a "modified wavelet transform" which improves resolution both in time and frequency. A simulation study illustrates the artifacts of every time-frequency representation (TFR) on pure sinusoids and gives performance evaluation of the different methods when searching a sinusoid embedded in a QRS complex. Analyses of real signals from healthy and pathological subjects confirm the simulation results and complete the characterization of ventricular late potentials (VLP) yet detected by signal averaging.

Automatic detection of wave boundaries in multilead ECG signals: validation with the CSE database

This paper presents an algorithm for automatically locating the waveform boundaries (the onsets and ends of P, QRS, and T waves) in multilead ECG signals (the 12 standard leads and the orthogonal XYZ leads). Given these locations, features of clinical importance (such as the RR interval, the PQ interval, the QRS duration, the ST segment, and the QT interval) may be measured readily. First, a multilead QRS detector locates each beat, using a differentiated and low-pass filtered ECG signal as input. Next, the waveform boundaries are located in each lead. The leads in which the detected electrical activity is of longest duration are used for the final determination of the waveform boundaries. The performance of our algorithm has been evaluated using the CSE multilead measurement database. In comparison with other algorithms tested by the CSE, our algorithm achieves better agreement with manual measurements of the T-wave end and of interval values, while its measurements of other waveform boundaries are within the range of the algorithm and manual measurements obtained by the CSE.

Acoustic analysis of vowel emission in obstructive sleep apnea

We studied vocalization in 18 men with obstructive sleep apnea syndrome (OSAS) (age, 49 [7.5] years; body mass index [BMI] 33.6 [7.6]) and 10 normal men as a control group (age, 46.7 [6.2] years; BMI 24.6 [2.2]). Polysomnographic data for patients with OSAS were as follows: total sleep time (TST), 387.5 [27.9] min; awake, 17.6 (12.6% TST); stage 1, 19.8 (18.7 percent TST); stage 2, 54.8 (23.2 percent TST); stage 3 and 4, 1.5 (0.3 percent TST); and stage REM, 4.2 (1.7 percent TST). Apnea hypopnea index (AHI) was 43.0 (18.2) and lowest O2 saturation was 73.6 (11.4). We recorded the following sounds in all subjects: /a/ as in "father"; /e/ as in "get"; /i/ as in "see"; /o/ as in "go"; /u/ as in "too." Three maneuvers for each vowel sound were taken for analysis. Signals were digitized at 10,000 Hz. Fast Fourier transformation was applied to segments of 512 points of each utterance corresponding to the vowel sound. The following parameters were obtained: maximum frequency of harmonics, mean frequency of harmonics, and the number of harmonics.

RESULTS

There were significant differences between both groups in the maximum frequency of harmonics of /i/ and /e/ vowels. (For /i/: 2,650 [672] Hz controls; 425 [71.2] Hz OSAS. For /e/: 2,605 [772.3] Hz controls; 1,250.0 [828.4] OSAS). The number of harmonics for /i/ vowel was 4.5 (1.2) for controls as compared with 2.7 (1) Hz for OSAS.

CONCLUSIONS

Vocalization in patients with OSAS is different from normal subjects. Vowel /i/ can distinguish these patients from normal subjects.

Orthonormal (Fourier and Walsh) models of time-varying evoked potentials in neurological injury

Estimation of time-varying changes in evoked potentials (EP's) has important applications, such as monitoring high-risk neurosurgical procedures. We test the hypothesis that injury related changes in EP signals may be modeled by orthonormal basis functions. We evaluate two models of time-varying EP signals: the Fourier series model (FSM) and the Walsh function model (WFM). We estimate the Fourier and Walsh coefficients with the aid of an adaptive least-mean-squares technique. Results from computer simulations illustrate how selection of model order and of the adaptation rate of the estimator affect the signal-to-noise ratio (SNR). The FSM results in a somewhat higher steady-state SNR than does the WFM; however, the WFM is less computationally complex than is the FSM. We apply these two orthonormal functions to evaluate transient response to hypoxic hypoxia in anesthetized cats. Trends of the first five frequencies (Fourier) and sequencies (Walsh) show that the lower frequencies and sequencies may be sensitive indicators of hypoxic neurological injury.

Adaptive filter for event-related bioelectric signals using an impulse correlated reference input: comparison with signal averaging techniques

Many bioelectric signals result from the electrical response of physiological systems to an impulse that can be internal (ECG signals) or external (evoked potentials). In this paper an adaptive impulse correlated filter (AICF) for event-related signals that are time-locked to a stimulus is presented. This filter estimates the deterministic component of the signal and removes the noise uncorrelated with the stimulus, even if this noise is colored, as in the case of evoked potentials. The filter needs two inputs: the signal (primary input) and an impulse correlated with the deterministic component (reference input). We use the LMS algorithm to adjust the weights in the adaptive process. First, we show that the AICF is equivalent to exponentially weighted averaging (EWA) when using the LMS algorithm. A quantitative analysis of the signal-to-noise ratio improvement, convergence, and misadjustment error is presented. A comparison of the AICF with ensemble averaging (EA) and moving window averaging (MWA) techniques is also presented. The adaptive filter is applied to real high-resolution ECG signals and time-varying somatosensory evoked potentials.