Detrended fluctuation analysis of EEG in sleep apnea using MIT/BIH polysomnography data

Classification of Sleep Quality and Aging as a Function of Brain Complexity: A Multiband Non-Linear EEG Analysis

Understanding and classifying brain states as a function of sleep quality and age has important implications for developing lifestyle-based interventions involving sleep hygiene. Current studies use an algorithm that captures non-linear features of brain complexity to differentiate awake electroencephalography (EEG) states, as a function of age and sleep quality. Fifty-eight participants were assessed using the Pittsburgh Sleep Quality Inventory (PSQI) and awake resting state EEG. Groups were formed based on age and sleep quality (younger adults n = 24, mean age = 24.7 years, SD = 3.43, good sleepers n = 11; older adults n = 34, mean age = 72.87; SD = 4.18, good sleepers n = 9). Ten non-linear features were extracted from multiband EEG analysis to feed several classifiers followed by a leave-one-out cross-validation. Brain state complexity accurately predicted (i) age in good sleepers, with 75% mean accuracy (across all channels) for lower frequencies (alpha, theta, and delta) and 95% accuracy at specific channels (temporal, parietal); and (ii) sleep quality in older groups with moderate accuracy (70 and 72%) across sub-bands with some regions showing greater differences. It also differentiated younger good sleepers from older poor sleepers with 85% mean accuracy across all sub-bands, and 92% at specific channels. Lower accuracy levels (<50%) were achieved in predicting sleep quality in younger adults. The algorithm discriminated older vs. younger groups excellently and could be used to explore intragroup differences in older adults to predict sleep intervention efficiency depending on their brain complexity.

Detection of Atrial Fibrillation Episodes based on 3D Algebraic Relationships between Cardiac Intervals

In this study, the notion of perfect matrices of Lagrange differences is employed to detect atrial fibrillation episodes based on three ECG parameters (JT interval, QRS interval, RR interval). The case study comprised 8 healthy individuals and 7 unhealthy individuals, and the mean and standard deviation of age was 65.84 ± 1.4 years, height was 1.75 ± 0.12 m, and weight was 79.4 ± 0.9 kg. Initially, it was demonstrated that the sensitivity of algebraic relationships between cardiac intervals increases when the dimension of the perfect matrices of Lagrange differences is extended from two to three. The baseline dataset was established using statistical algorithms for classification by means of the developed decision support system. The classification helps to determine whether the new incoming candidate has indications of atrial fibrillation or not. The application of probability distribution graphs and semi-gauge indicator techniques aided in visualizing the categorization of the new candidates. Though the study's data are limited, this work provides a strong foundation for (1) validating the sensitivity of the perfect matrices of Lagrange differences, (2) establishing a robust baseline dataset for supervised classification, and (3) classifying new incoming candidates within the classification framework. From a clinical standpoint, the developed approach assists in the early detection of atrial fibrillation in an individual.

COVID-19 activity screening by a smart-data-driven multi-band voice analysis

COVID-19 is a disease caused by the new coronavirus SARS-COV-2 which can lead to severe respiratory infections. Since its first detection it caused more than six million worldwide deaths. COVID-19 diagnosis non-invasive and low-cost methods with faster and accurate results are still needed for a fast disease control. In this research, 3 different signal analyses have been applied (per broadband, per sub-bands and per broadband & sub-bands) to Cough, Breathing & Speech signals of Coswara dataset to extract non-linear patterns (Energy, Entropies, Correlation Dimension, Detrended Fluctuation Analysis, Lyapunov Exponent & Fractal Dimensions) for feeding a XGBoost classifier to discriminate COVID-19 activity on its different stages. Classification accuracies ranged between 83.33% and 98.46% have been achieved, surpassing the state-of-art methods in some comparisons. It should be empathized the 98.46% of accuracy reached on pair Healthy Controls vs all COVID-19 stages. The results shows that the method may be adequate for COVID-19 diagnosis screening assistance.

Decreased sample entropy during sleep-to-wake transition in sleep apnea patients

Objective. This study aimed to prove that there is a sudden change in the human physiology system when switching from one sleep stage to another and physical threshold-based sample entropy (SampEn) is able to capture this transition in an RR interval time series from patients with disorders such as sleep apnea.Approach. Physical threshold-based SampEn was used to analyze different sleep-stage RR segments from sleep apnea subjects in the St. Vincents University Hospital/University College Dublin Sleep Apnea Database, and SampEn differences were compared between two consecutive sleep stages. Additionally, other standard heart rate variability (HRV) measures were also analyzed to make comparisons.Main results. The findings suggested that the sleep-to-wake transitions presented a SampEn decrease significantly larger than intra-sleep ones (P < 0.01), which outperformed other standard HRV measures. Moreover, significant entropy differences between sleep and subsequent wakefulness appeared when the previous sleep stage was either S1 (P < 0.05), S2 (P < 0.01) or S4 (P < 0.05).Significance. The results demonstrated that physical threshold-based SampEn has the capability of depicting physiological changes in the cardiovascular system during the sleep-to-wake transition in sleep apnea patients and it is more reliable than the other analyzed HRV measures. This noninvasive HRV measure is a potential tool for further evaluation of sleep physiological time series.

Detrended fluctuation analysis in the presurgical evaluation of parietal lobe epilepsy patients

OBJECTIVE

To examine the usability of long-range temporal correlations (LRTCs) in non-invasive localization of the epileptogenic zone (EZ) in refractory parietal lobe epilepsy (RPLE) patients.

METHODS

We analyzed 10 RPLE patients who had presurgical MEG and underwent epilepsy surgery. We quantified LRTCs with detrended fluctuation analysis (DFA) at four frequency bands for 200 cortical regions estimated using individual source models. We correlated individually the DFA maps to the distance from the resection area and from cortical locations of interictal epileptiform discharges (IEDs). Additionally, three clinical experts inspected the DFA maps to visually assess the most likely EZ locations.

RESULTS

The DFA maps correlated with the distance to resection area in patients with type II focal cortical dysplasia (FCD) (p<0.05), but not in other etiologies. Similarly, the DFA maps correlated with the IED locations only in the FCD II patients. Visual analysis of the DFA maps showed high interobserver agreement and accuracy in FCD patients in assigning the affected hemisphere and lobe.

CONCLUSIONS

Aberrant LRTCs correlate with the resection areas and IED locations.

SIGNIFICANCE

This methodological pilot study demonstrates the feasibility of approximating cortical LRTCs from MEG that may aid in the EZ localization and provide new non-invasive insight into the presurgical evaluation of epilepsy.

The Analysis of Electroencephalography Changes Before and After a Single Neurofeedback Alpha/Theta Training Session in University Students

The underlying mechanisms of alpha/theta neurofeedback training have not been fully determined. Therefore, this study aimed to test the changes in the brain state feedback during the alpha/theta training. Twenty-seven healthy participants were trained during a single session of the alpha/theta protocol, and the resting quantitative electroencephalography (QEEG) was assessed before and after training. QEEG was recorded at eight scalp locations (F3, F4, C3, C4, T3, T4, O1, and O2), and the absolute power, relative power, ratio of sensory-motor rhythm beta (SMR) to theta (RST), ratio of SMR-mid beta to theta (RSMT), ratio of mid beta to theta (RMT), ratio of alpha to high beta (RAHB), and scaling exponent of detrended fluctuation analysis by each band were measured. The results indicated a significant increase of absolute alpha power, especially the slow alpha band, at all electrodes except T3 and T4. Moreover, the relative alpha power, especially the slow alpha band, showed a significant increase at all electrodes. The relative theta power showed a significant decrease at all electrodes, except T3. A significant decrease in relative beta power, relative lower beta power and relative mid beta power was observed at O1. RST (at C4, O1, and O2), RSMT and RMT (at F4, C4, O1 and O2), and RAHB (at all electrodes) showed significant increase. Scaling exponents at all electrodes except T3 showed a significant decrease. These findings indicate that a one-time session of alpha/theta training might have the possibility to enhance both vigilance and concentration, thus stabilizing the overall brain function.

Joint Classification and Prediction CNN Framework for Automatic Sleep Stage Classification

Correctly identifying sleep stages is important in diagnosing and treating sleep disorders. This paper proposes a joint classification-and-prediction framework based on convolutional neural networks (CNNs) for automatic sleep staging, and, subsequently, introduces a simple yet efficient CNN architecture to power the framework. Given a single input epoch, the novel framework jointly determines its label (classification) and its neighboring epochs' labels (prediction) in the contextual output. While the proposed framework is orthogonal to the widely adopted classification schemes, which take one or multiple epochs as contextual inputs and produce a single classification decision on the target epoch, we demonstrate its advantages in several ways. First, it leverages the dependency among consecutive sleep epochs while surpassing the problems experienced with the common classification schemes. Second, even with a single model, the framework has the capacity to produce multiple decisions, which are essential in obtaining a good performance as in ensemble-of-models methods, with very little induced computational overhead. Probabilistic aggregation techniques are then proposed to leverage the availability of multiple decisions. To illustrate the efficacy of the proposed framework, we conducted experiments on two public datasets: Sleep-EDF Expanded (Sleep-EDF), which consists of 20 subjects, and Montreal Archive of Sleep Studies (MASS) dataset, which consists of 200 subjects. The proposed framework yields an overall classification accuracy of 82.3% and 83.6%, respectively. We also show that the proposed framework not only is superior to the baselines based on the common classification schemes but also outperforms existing deep-learning approaches. To our knowledge, this is the first work going beyond the standard single-output classification to consider multitask neural networks for automatic sleep staging. This framework provides avenues for further studies of different neural-network architectures for automatic sleep staging.

Robust Fetal Heart Beat Detection via R-Peak Intervals Distribution

Monitoring fetal heart rate during pregnancy is essential to assist clinicians in making more timely decisions. Non-invasive monitoring of fetal heart activities using abdominal ECGs is useful for diagnosis of heart defects. However, the extracted fetal ECGs are usually too weak to be robustly detected. Thus, it is a necessity to enhance fetal R-peak since their peaks may be hidden within the signal due to the immaturity of the fetal cardiovascular system. Therefore, to improve the detection of the fetal heartbeat, a novel fetal R-peak enhancement technique was proposed to statistically generate the weighting mask according to the distribution of the neighboring temporal intervals between each pair of peaks. Two sets of simulations were designed to validate the reliability of the method: challenges with different levels of (1) noise contamination and (2) R-peak interval changing rate. The simulation results showed that the weighting mask improved the accuracy of the R-peak detection rate by 25% and decreased the false alarm rate by 20% with white noise contamination, and ensured high R-peak detection rate (>80%), especially with mild noise contamination (noise amplitude ratio <1.5 and noise rate per minute <25%). For the simulations with continuous R-peak intervals changing, the masking process can still effectively eliminate noise contamination especially when the amplitude of the sinusoidal fetal R-R intervals is lower than 50 ms. For the real fetus ECGs, the detection rate was increased by 3.498%, whereas the false alarm rate was decreased by 3.933%. Next, we implemented the fetal R-peak enhancement technique to investigate fractal regulation and multiscale entropy of the real fetal heartbeat intervals. Both scaling exponent (∼0.6 to ∼1 in scale 4-15) and entropy measure (scale 6-10) increased with gestational ages (22-40 weeks). The results confirmed fractal slope and complexity of fetal heartbeat intervals can reflect the maturation of fetus organism.

Detrended fluctuation analysis of the oximetry signal to assist in paediatric sleep apnoea-hypopnoea syndrome diagnosis

OBJECTIVE

To evaluate whether detrended fluctuation analysis (DFA) provides information that improves the diagnostic ability of the oximetry signal in the diagnosis of paediatric sleep apnoea-hypopnoea syndrome (SAHS).

APPROACH

A database composed of 981 blood oxygen saturation (SpO₂) recordings in children was used to extract DFA-derived features in order to quantify the scaling behaviour and the fluctuations of the SpO₂ signal. The 3% oxygen desaturation index (ODI3) was also computed for each subject. Fast correlation-based filter (FCBF) was then applied to select an optimum subset of relevant and non-redundant features. This subset fed a multi-layer perceptron (MLP) neural network to estimate the apnoea-hypopnoea index (AHI).

MAIN RESULTS

ODI3 and four features from the DFA reached significant differences associated with the severity of SAHS. An optimum subset composed of the slope in the first scaling region of the DFA profile and the ODI3 was selected using FCBF applied to the training set (60% of samples). The MLP model trained with this feature subset showed good agreement with the actual AHI, reaching an intra-class correlation coefficient of 0.891 in the test set (40% of samples). Furthermore, the estimated AHI showed high diagnostic ability, reaching an accuracy of 82.7%, 81.9%, and 91.1% using three common AHI cut-offs of 1, 5, and 10 events per hour (e h^-1), respectively. These results outperformed the overall performance of ODI3.

SIGNIFICANCE

DFA may serve as a reliable tool to improve the diagnostic performance of oximetry recordings in the evaluation of paediatric patients with symptoms suggestive of SAHS.

The long-range correlation and evolution law of centennial-scale temperatures in Northeast China

This paper applies the detrended fluctuation analysis (DFA) method to investigate the long-range correlation of monthly mean temperatures from three typical measurement stations at Harbin, Changchun, and Shenyang in Northeast China from 1909 to 2014. The results reveal the memory characteristics of the climate system in this region. By comparing the temperatures from different time periods and investigating the variations of its scaling exponents at the three stations during these different time periods, we found that the monthly mean temperature has long-range correlation, which indicates that the temperature in Northeast China has long-term memory and good predictability. The monthly time series of temperatures over the past 106 years also shows good long-range correlation characteristics. These characteristics are also obviously observed in the annual mean temperature time series. Finally, we separated the centennial-length temperature time series into two time periods. These results reveal that the long-range correlations at the Harbin station over these two time periods have large variations, whereas no obvious variations are observed at the other two stations. This indicates that warming affects the regional climate system’s predictability differently at different time periods. The research results can provide a quantitative reference point for regional climate predictability assessment and future climate model evaluation.

Long-range temporal correlations in the brain distinguish conscious wakefulness from induced unconsciousness

Rhythmic neuronal synchronization across large-scale networks is thought to play a key role in the regulation of conscious states. Changes in neuronal oscillation amplitude across states of consciousness have been widely reported, but little is known about possible changes in the temporal dynamics of these oscillations. The temporal structure of brain oscillations may provide novel insights into the neural mechanisms underlying consciousness. To address this question, we examined long-range temporal correlations (LRTC) of EEG oscillation amplitudes recorded during both wakefulness and anesthetic-induced unconsciousness. Importantly, the time-varying EEG oscillation envelopes were assessed over the course of a sevoflurane sedation protocol during which the participants alternated between states of consciousness and unconsciousness. Both spectral power and LRTC in oscillation amplitude were computed across multiple frequency bands. State-dependent differences in these features were assessed using non-parametric tests and supervised machine learning. We found that periods of unconsciousness were associated with increases in LRTC in beta (15-30Hz) amplitude over frontocentral channels and with a suppression of alpha (8-13Hz) amplitude over occipitoparietal electrodes. Moreover, classifiers trained to predict states of consciousness on single epochs demonstrated that the combination of beta LRTC with alpha amplitude provided the highest classification accuracy (above 80%). These results suggest that loss of consciousness is accompanied by an augmentation of temporal persistence in neuronal oscillation amplitude, which may reflect an increase in regularity and a decrease in network repertoire compared to the brain's activity during resting-state consciousness.

Auto-correlation in the motor/imaginary human EEG signals: A vision about the FDFA fluctuations

In this paper we analyzed, by the F_DFA root mean square fluctuation (rms) function, the motor/imaginary human activity produced by a 64-channel electroencephalography (EEG). We utilized the Physionet on-line databank, a publicly available database of human EEG signals, as a standardized reference database for this study. Herein, we report the use of detrended fluctuation analysis (DFA) method for EEG analysis. We show that the complex time series of the EEG exhibits characteristic fluctuations depending on the analyzed channel in the scalp-recorded EEG. In order to demonstrate the effectiveness of the proposed technique, we analyzed four distinct channels represented here by F₃32, F₆37 (frontal region of the head) and P₃49, P₆54 (parietal region of the head). We verified that the amplitude of the F_DFA rms function is greater for the frontal channels than for the parietal. To tabulate this information in a better way, we define and calculate the difference between F_DFA (in log scale) for the channels, thus defining a new path for analysis of EEG signals. Finally, related to the studied EEG signals, we obtain the auto-correlation exponent, α_DFA by DFA method, that reveals self-affinity at specific time scale. Our results shows that this strategy can be applied to study the human brain activity in EEG processing.

Detrended fluctuation analysis for major depressive disorder

Clinical utility of Electroencephalography (EEG) based diagnostic studies is less clear for major depressive disorder (MDD). In this paper, a novel machine learning (ML) scheme was presented to discriminate the MDD patients and healthy controls. The proposed method inherently involved feature extraction, selection, classification and validation. The EEG data acquisition involved eyes closed (EC) and eyes open (EO) conditions. At feature extraction stage, the de-trended fluctuation analysis (DFA) was performed, based on the EEG data, to achieve scaling exponents. The DFA was performed to analyzes the presence or absence of long-range temporal correlations (LRTC) in the recorded EEG data. The scaling exponents were used as input features to our proposed system. At feature selection stage, 3 different techniques were used for comparison purposes. Logistic regression (LR) classifier was employed. The method was validated by a 10-fold cross-validation. As results, we have observed that the effect of 3 different reference montages on the computed features. The proposed method employed 3 different types of feature selection techniques for comparison purposes as well. The results show that the DFA analysis performed better in LE data compared with the IR and AR data. In addition, during Wilcoxon ranking, the AR performed better than LE and IR. Based on the results, it was concluded that the DFA provided useful information to discriminate the MDD patients and with further validation can be employed in clinics for diagnosis of MDD.

Generalized Information Equilibrium Approaches to EEG Sleep Stage Discrimination

Recent advances in neuroscience have raised the hypothesis that the underlying pattern of neuronal activation which results in electroencephalography (EEG) signals is via power-law distributed neuronal avalanches, while EEG signals are nonstationary. Therefore, spectral analysis of EEG may miss many properties inherent in such signals. A complete understanding of such dynamical systems requires knowledge of the underlying nonequilibrium thermodynamics. In recent work by Fielitz and Borchardt (2011, 2014), the concept of information equilibrium (IE) in information transfer processes has successfully characterized many different systems far from thermodynamic equilibrium. We utilized a publicly available database of polysomnogram EEG data from fourteen subjects with eight different one-minute tracings of sleep stage 2 and waking and an overlapping set of eleven subjects with eight different one-minute tracings of sleep stage 3. We applied principles of IE to model EEG as a system that transfers (equilibrates) information from the time domain to scalp-recorded voltages. We find that waking consciousness is readily distinguished from sleep stages 2 and 3 by several differences in mean information transfer constants. Principles of IE applied to EEG may therefore prove to be useful in the study of changes in brain function more generally.

Bach Is the Father of Harmony: Revealed by a 1/f Fluctuation Analysis across Musical Genres

Harmony is a fundamental attribute of music. Close connections exist between music and mathematics since both pursue harmony and unity. In music, the consonance of notes played simultaneously partly determines our perception of harmony; associates with aesthetic responses; and influences the emotion expression. The consonance could be considered as a window to understand and analyze harmony. Here for the first time we used a 1/f fluctuation analysis to investigate whether the consonance fluctuation structure in music with a wide range of composers and genres followed the scale free pattern that has been found for pitch, melody, rhythm, human body movements, brain activity, natural images and geographical features. We then used a network graph approach to investigate which composers were the most influential both within and across genres. Our results showed that patterns of consonance in music did follow scale-free characteristics, suggesting that this feature is a universally evolved one in both music and the living world. Furthermore, our network analysis revealed that Bach’s harmony patterns were having the most influence on those used by other composers, followed closely by Mozart.

Exploring strategies for classification of external stimuli using statistical features of the plant electrical response

Plants sense their environment by producing electrical signals which in essence represent changes in underlying physiological processes. These electrical signals, when monitored, show both stochastic and deterministic dynamics. In this paper, we compute 11 statistical features from the raw non-stationary plant electrical signal time series to classify the stimulus applied (causing the electrical signal). By using different discriminant analysis-based classification techniques, we successfully establish that there is enough information in the raw electrical signal to classify the stimuli. In the process, we also propose two standard features which consistently give good classification results for three types of stimuli--sodium chloride (NaCl), sulfuric acid (H₂SO₄) and ozone (O₃). This may facilitate reduction in the complexity involved in computing all the features for online classification of similar external stimuli in future.

Computer-Aided Diagnosis of Depression Using EEG Signals

The complex, nonlinear and non-stationary electroencephalogram (EEG) signals are very tedious to interpret visually and highly difficult to extract the significant features from them. The linear and nonlinear methods are effective in identifying the changes in EEG signals for the detection of depression. Linear methods do not exhibit the complex dynamical variations in the EEG signals. Hence, chaos theory and nonlinear dynamic methods are widely used in extracting the EEG signal features for computer-aided diagnosis (CAD) of depression. Hence, this article presents the recent efforts on CAD of depression using EEG signals with a focus on using nonlinear methods. Such a CAD system is simple to use and may be used by the clinicians as a tool to confirm their diagnosis. It should be of a particular value to enable the early detection of depression.

Automatic detection of sleep apnea based on EEG detrended fluctuation analysis and support vector machine

Sleep apnea syndrome (SAS) is prevalent in individuals and recently, there are many studies focus on using simple and efficient methods for SAS detection instead of polysomnography. However, not much work has been done on using nonlinear behavior of the electroencephalogram (EEG) signals. The purpose of this study is to find a novel and simpler method for detecting apnea patients and to quantify nonlinear characteristics of the sleep apnea. 30 min EEG scaling exponents that quantify power-law correlations were computed using detrended fluctuation analysis (DFA) and compared between six SAS and six healthy subjects during sleep. The mean scaling exponents were calculated every 30 s and 360 control values and 360 apnea values were obtained. These values were compared between the two groups and support vector machine (SVM) was used to classify apnea patients. Significant difference was found between EEG scaling exponents of the two groups (p < 0.001). SVM was used and obtained high and consistent recognition rate: average classification accuracy reached 95.1% corresponding to the sensitivity 93.2% and specificity 98.6%. DFA of EEG is an efficient and practicable method and is helpful clinically in diagnosis of sleep apnea.

Quantitative evaluation of the use of actigraphy for neurological and psychiatric disorders

Quantitative and objective evaluation of disease severity and/or drug effect is necessary in clinical practice. Wearable accelerometers such as an actigraph enable long-term recording of a patient's movement during activities and they can be used for quantitative assessment of symptoms due to various diseases. We reviewed some applications of actigraphy with analytical methods that are sufficiently sensitive and reliable to determine the severity of diseases and disorders such as motor and nonmotor disorders like Parkinson's disease, sleep disorders, depression, behavioral and psychological symptoms of dementia (BPSD) for vascular dementia (VD), seasonal affective disorder (SAD), and stroke, as well as the effects of drugs used to treat them. We believe it is possible to develop analytical methods to assess more neurological or psychopathic disorders using actigraphy records.

A new algorithm for quadratic sample entropy optimization for very short biomedical signals: application to blood pressure records

This paper describes a new method to optimize the computation of the quadratic sample entropy (QSE) metric. The objective is to enhance its segmentation capability between pathological and healthy subjects for short and unevenly sampled biomedical records, like those obtained using ambulatory blood pressure monitoring (ABPM). In ABPM, blood pressure is measured every 20-30 min during 24h while patients undergo normal daily activities. ABPM is indicated for a number of applications such as white-coat, suspected, borderline, or masked hypertension. Hypertension is a very important clinical issue that can lead to serious health implications, and therefore its identification and characterization is of paramount importance. Nonlinear processing of signals by means of entropy calculation algorithms has been used in many medical applications to distinguish among signal classes. However, most of these methods do not perform well if the records are not long enough and/or not uniformly sampled. That is the case for ABPM records. These signals are extremely short and scattered with outliers or missing/resampled data. This is why ABPM Blood pressure signal screening using nonlinear methods is a quite unexplored field. We propose an additional stage for the computation of QSE independently of its parameter r and the input signal length. This enabled us to apply a segmentation process to ABPM records successfully. The experimental dataset consisted of 61 blood pressure data records of control and pathological subjects with only 52 samples per time series. The entropy estimation values obtained led to the segmentation of the two groups, while other standard nonlinear methods failed.

Multifractal detrended fluctuation analysis of human EEG: preliminary investigation and comparison with the wavelet transform modulus maxima technique

Recently, many lines of investigation in neuroscience and statistical physics have converged to raise the hypothesis that the underlying pattern of neuronal activation which results in electroencephalography (EEG) signals is nonlinear, with self-affine dynamics, while scalp-recorded EEG signals themselves are nonstationary. Therefore, traditional methods of EEG analysis may miss many properties inherent in such signals. Similarly, fractal analysis of EEG signals has shown scaling behaviors that may not be consistent with pure monofractal processes. In this study, we hypothesized that scalp-recorded human EEG signals may be better modeled as an underlying multifractal process. We utilized the Physionet online database, a publicly available database of human EEG signals as a standardized reference database for this study. Herein, we report the use of multifractal detrended fluctuation analysis on human EEG signals derived from waking and different sleep stages, and show evidence that supports the use of multifractal methods. Next, we compare multifractal detrended fluctuation analysis to a previously published multifractal technique, wavelet transform modulus maxima, using EEG signals from waking and sleep, and demonstrate that multifractal detrended fluctuation analysis has lower indices of variability. Finally, we report a preliminary investigation into the use of multifractal detrended fluctuation analysis as a pattern classification technique on human EEG signals from waking and different sleep stages, and demonstrate its potential utility for automatic classification of different states of consciousness. Therefore, multifractal detrended fluctuation analysis may be a useful pattern classification technique to distinguish among different states of brain function.

Two-dimensional matrix algorithm using detrended fluctuation analysis to distinguish Burkitt and diffuse large B-cell lymphoma

A detrended fluctuation analysis (DFA) method is applied to image analysis. The 2-dimensional (2D) DFA algorithms is proposed for recharacterizing images of lymph sections. Due to Burkitt lymphoma (BL) and diffuse large B-cell lymphoma (DLBCL), there is a significant different 5-year survival rates after multiagent chemotherapy. Therefore, distinguishing the difference between BL and DLBCL is very important. In this study, eighteen BL images were classified as group A, which have one to five cytogenetic changes. Ten BL images were classified as group B, which have more than five cytogenetic changes. Both groups A and B BLs are aggressive lymphomas, which grow very fast and require more intensive chemotherapy. Finally, ten DLBCL images were classified as group C. The short-term correlation exponent α1 values of DFA of groups A, B, and C were 0.370 ± 0.033, 0.382 ± 0.022, and 0.435 ± 0.053, respectively. It was found that α1 value of BL image was significantly lower (P < 0.05) than DLBCL. However, there is no difference between the groups A and B BLs. Hence, it can be concluded that α1 value based on DFA statistics concept can clearly distinguish BL and DLBCL image.

An ensemble system for automatic sleep stage classification using single channel EEG signal

The present work aims at automatic identification of various sleep stages like, sleep stages 1, 2, slow wave sleep (sleep stages 3 and 4), REM sleep and wakefulness from single channel EEG signal. Automatic scoring of sleep stages was performed with the help of pattern recognition technique which involves feature extraction, selection and finally classification. Total 39 numbers of features from time domain, frequency domain and from non-linear analysis were extracted. After extraction of features, SVM based recursive feature elimination (RFE) technique was used to find the optimum number of feature subset which can provide significant classification performance with reduced number of features for the five different sleep stages. Finally for classification, binary SVMs were combined with one-against-all (OAA) strategy. Careful extraction and selection of optimum feature subset helped to reduce the classification error to 8.9% for training dataset, validated by k-fold cross-validation (CV) technique and 10.61% in the case of independent testing dataset. Agreement of the estimated sleep stages with those obtained by expert scoring for all sleep stages of training dataset was 0.877 and for independent testing dataset it was 0.8572. The proposed ensemble SVM-based method could be used as an efficient and cost-effective method for sleep staging with the advantage of reducing stress and burden imposed on subjects.

DETRENDED FLUCTUATION ANALYSES OF SHORT-TERM HEART RATE VARIABILITY IN SURGICAL INTENSIVE CARE UNITS

We examine the dynamics of complex physiologic fluctuations using methods developed very recently in statistical physics. The method based on detrended fluctuation analysis (DFA) has been used to investigate the profile of different types of physiologic states under long term (i.e., 24 hr) analysis of heart rate variability (HRV). In this paper, this method to investigate the output of central physiologic control system under short term (i.e., 1 hr) of HRV in surgical intensive care units (SICU). Electrocardiograph (ECG) signals lasting around 1 hr were collected from ten college student volunteers as group A. Ten computes-generates white noise signals as group B also provided ECG signals lasting around 1 hr. Finally, seventeen patients representing 37 cases undergoing different types of neurosurgery were studied as group C. From this group, 25 cases were selected from 15 patients with brain injury and 12 cases were selected from 2 patients with septicemia. Group A and B were used as high and low limits of baseline for comparison with pathologic states in the SICU. The a values of DFA of groups A, B, and C were 0.958 ± 0.034, 0.521 ± 0.010, and 0.815 ± 0.183, respectively. It was found that the α value of patients in the SICU was significantly lower (P < 0.05) than that of healthy volunteers and significantly higher (P < 0.05) than white noise signals. Hence, it can be concluded that α values based on the DFA statistical concept can clearly distinguish pathologic states in SICU patients from the healthy group and from white noise signals.

Fractal analysis of EEG upon auditory stimulation during waking and hypnosis in healthy volunteers

The authors tested fluctuation analyses (DFA) of EEGs upon auditory stimulation in waking and hypnotic states as related to topography and hypnotizability. They administered the Hypnotic Induction Profile (HIP), Dissociation Experience Scale, and Tellegen Absorption Scale to 10 healthy volunteers and measured subjects' EEGs while the subjects listened to sounds, either selecting or ignoring tones of different decibels, in waking and hypnotic states. DFA scaling exponents were closest to 0.5 when subjects reported the tones in the hypnotic state. Different DFA values at C3 showed significant positive correlations with the HIP eye-roll sign. Adding to the literature supporting the state theory of hypnosis, the DFA values at F3 and C3 showed significant differences between waking and hypnotic states. Application of auditory stimuli is useful for understanding neurophysiological characteristics of hypnosis using DFA.

Long-range correlations of different EEG derivations in rats: sleep stage-dependent generators may play a key role

For the electroencephalogram (EEG), topographic differences in the long-range temporal correlations would imply that these signals might be affected by specific mechanisms related to the generation of a given neuronal process. So the properties of the generators of various EEG oscillations might be investigated by their spatial differences of the long-range temporal correlations. In the present study, these correlations were characterized with respect to their topography during different vigilance states by detrended fluctuation analysis (DFA). The results indicated that (1) most of the scaling exponents acquired from different EEG derivations for various oscillations were significantly different in each vigilance state; these differences might be resulted from the different quantities and different locations of sleep stage-dependent generators of various neuronal processes; (2) there might be multiple generators of delta and theta over the brain and many of them were sleep stage-dependent; (3) the best site of the frontal electrode in a fronto-parietal bipolar electrode for sleep staging might be above the anterior midline cortex. We suggest that DFA analysis can be used to explore the properties of the generators of a given neuronal oscillation, and the localizations of these generators if more electrodes are involved.

Dynamics of electroencephalogram entropy and pitfalls of scaling detection

In recent studies a number of research groups have determined that human electroencephalograms (EEG) have scaling properties. In particular, a crossover between two regions with different scaling exponents has been reported. Herein we study the time evolution of diffusion entropy to elucidate the scaling of EEG time series. For a cohort of 20 awake healthy volunteers with closed eyes, we find that the diffusion entropy of EEG increments (obtained from EEG waveforms by differencing) exhibits three features: short-time growth, an alpha wave related oscillation whose amplitude gradually decays in time, and asymptotic saturation which is achieved after approximately 1 s. This analysis suggests a linear, stochastic Ornstein-Uhlenbeck Langevin equation with a quasiperiodic forcing (whose frequency and/or amplitude may vary in time) as the model for the underlying dynamics. This model captures the salient properties of EEG dynamics. In particular, both the experimental and simulated EEG time series exhibit short-time scaling which is broken by a strong periodic component, such as alpha waves. The saturation of EEG diffusion entropy precludes the existence of asymptotic scaling. We find that the crossover between two scaling regions seen in detrended fluctuation analysis (DFA) of EEG increments does not originate from the underlying dynamics but is merely an artifact of the algorithm. This artifact is rooted in the failure of the "trend plus signal" paradigm of DFA.

Quantification of brain macrostates using dynamical nonstationarity of physiological time series

The brain shows complex, nonstationarity temporal dynamics, with abrupt micro- and macrostate transitions during its information processing. Detecting and characterizing these transitions in dynamical states of the brain is a critical issue in the field of neuroscience and psychiatry. In the current study, a novel method is proposed to quantify brain macrostates (e.g., sleep stages or cognitive states) from shifts of dynamical microstates or dynamical nonstationarity. A ``dynamical microstate'' is a temporal unit of the information processing in the brain with fixed dynamical parameters and specific spatial distribution. In this proposed approach, a phase-space-based dynamical dissimilarity map (DDM) is used to detect transitions between dynamically stationary microstates in the time series, and Tsallis time-dependent entropy is applied to quantify dynamical patterns of transitions in the DDM. We demonstrate that the DDM successfully detects transitions between microstates of different temporal dynamics in the simulated physiological time series against high levels of noise. Based on the assumption of nonlinear, deterministic brain dynamics, we also demonstrate that dynamical nonstationarity analysis is useful to quantify brain macrostates (sleep stages I, II, III, IV, and rapid eye movement (REM) sleep) from sleep EEGs with an overall accuracy of 77%. We suggest that dynamical nonstationarity is a useful tool to quantify macroscopic mental states (statistical integration) of the brain using dynamical transitions at the microscopic scale in physiological data.

Quantitative study of the sleep onset period via detrended fluctuation analysis: normal vs. narcoleptic subjects

OBJECTIVE

To examine the process of the sleep onset quantitatively and explore differences between narcoleptics and controls during the sleep onset period (SOP).

METHOD

Dynamic detrended fluctuation analysis (DFA) was applied to electroencephalograms recorded during multiple sleep latency tests of 11 drug-free narcoleptic patients (19.3+/-4.4 yrs; 8 males) and 9 healthy controls (23.8+/-6.3 yrs; 6 males). The SOP of each group was estimated by fitting the time courses of the DFA scaling exponents to a parametric curve.

RESULTS

The sequence of DFA exponents showed that electrophysiological brain activity was changing rapidly across the SOP. This transition was also verified by a conventional method (i.e., dynamic spectral analysis). The SOP durations of narcoleptics and controls were estimated as 239+/-25 s and 145+/-20 s, respectively.

CONCLUSIONS

The significantly larger SOP of narcoleptics, compared to controls, is consistent with the wake state of narcolepsy being more susceptible to sleep due to a lower barrier to transitioning to sleep.

SIGNIFICANCE

Our results suggest that electrophysiological signatures of narcolepsy could be quantified by dynamic DFA, so the method may have promise as a potential tool to help the diagnosis of narcolepsy despite the present study's limited sample size.

Cross-modulated amplitudes and frequencies characterize interacting components in complex systems

The dynamics of complex systems is characterized by oscillatory components on many time scales. To study the interactions between these components we analyze the cross modulation of their instantaneous amplitudes and frequencies, separating synchronous and antisynchronous modulation. We apply our novel technique to brain-wave oscillations in the human electroencephalogram and show that interactions between the alpha wave and the delta or beta wave oscillators as well as spatial interactions can be quantified and related with physiological conditions (e.g., sleep stages). Our approach overcomes the limitation to oscillations with similar frequencies and enables us to quantify directly nonlinear effects such as positive or negative frequency modulation.

Theoretical basis for identification of different anesthetic states based on routinely recorded EEG during operation

In this paper, we present a new method to identify anesthetic states based on routinely recorded electroencephalogram (EEG). The identification of anesthesia stages are conducted using fast Fourier transform (FFT) and modified detrended fluctuation analysis (DFA) method. Simulation results demonstrate that this new method can clearly discriminate all five anesthesia states: very deep anesthesia, deep anesthesia, moderate anesthesia, light anesthesia and awake.

Scale invariance of human electroencephalogram signals in sleep

In this paper, we investigate the dynamical properties of electroencephalogram (EEG) signals of humans in sleep. By using a modified random walk method, we demonstrate that scale-invariance is embedded in EEG signals after a detrending procedure is applied. Furthermore, we study the dynamical evolution of the probability density function (PDF) of the detrended EEG signals by nonextensive statistical modeling. It displays a scale-independent property, which is markedly different from the usual scale-dependent PDF evolution and cannot be described by the Fokker-Planck equation.

The modifications of the long-range temporal correlations of the sleep EEG due to major depressive episode disappear with the status of remission

OBJECTIVE

The aim of the present study is to investigate the scaling properties of the sleep electroencephalogram (EEG) in remitted depressed men, and to evaluate if a past history of major depressive disorder (MDD) could modify significantly and definitively, as a "scar marker," the dynamics of the sleep EEG time series.

METHODOLOGY

Whole night sleep electroencephalogram signals were recorded in 24 men: 10 untreated depressed men in full to partial remission (42.43+/-5.62 years) and 14 healthy subjects (42.8+/-8.55 years). Scaling properties in these time series were investigated with detrended fluctuation analysis (DFA) (time range: 0.16-2.00 s). The scaling exponent alpha was determined in stage 2, in slow wave sleep (stages 3 and 4), and during rapid eye movement (REM) sleep. Forty-five epochs of 20 s were chosen randomly in each of these stages for each subject in both groups.

RESULTS

We did not observe a significant difference and deviation of the scaling exponents between the two groups during the three sleep stages of interest.

CONCLUSION

In this study, we do not observe any functional sequelae of a past history of one or more unipolar major depressive episode on the fluctuation properties of the sleep EEG. This finding is a sign of similar underlying neuronal dynamics in healthy controls and patients with a lifetime history of MDD. This study gives an additional argument to the theory that depression does not modify definitively the dynamics of the neuronal networks and is therefore against the "depressive scar hypothesis," in which permanent residual deficit is created by the acute state of the depressive disease.

Detrended Fluctuation Analysis of EEG as a Measure of Depth of Anesthesia

For several decades, a number of methods have been developed for the noninvasive assessment of the level of consciousness during general anesthesia. In this paper, detrended fluctuation analysis is used to study the scaling behavior of the electroencephalogram as a measure of the level of consciousness. Three indexes are proposed in order to characterize the patient state. Statistical analysis demonstrates that they allow significant discrimination between the awake, sedated and anesthetized states. Two of them present a good correlation with established indexes of depth of anesthesia. The scaling behavior has been found related to the depth of anesthesia and the methodology allows real-time implementation, which enables its application in monitoring devices.

Feedback modulates the temporal scale-free dynamics of brain electrical activity in a hypothesis testing task

We used the electroencephalogram (EEG) to investigate whether positive and negative performance feedbacks exert different long-lasting modulations of electrical activity in a reasoning task. Nine college students serially tested hypotheses concerning a hidden rule by judging its presence or absence in triplets of digits, and revised them on the basis of an exogenous performance feedback. The scaling properties of the transition period between feedback and triplet presentation were investigated with detrended fluctuation analysis (DFA). DFA showed temporal scale-free dynamics of EEG activity in both feedback conditions for time scales larger than 150 ms. Furthermore, DFA revealed that negative feedback elicits significantly higher scaling exponents than positive feedback. This effect covers a wide network comprising parieto-occipital and left frontal regions. We thus showed that specific task demands can modify the temporal scale-free dynamics of the ongoing brain activity. Putative neural correlates of these long-lasting feedback-specific modulations are proposed.

Characterization of the sleep EEG in acutely depressed men using detrended fluctuation analysis

OBJECTIVE

The aim of the present paper is to study the fluctuations of the sleep EEG over various time scales during a specific pathological condition: major depressive episode. Focus is made on scaling behaviour, which is the signature of the absence of characteristic time scale, and the presence of long-range correlations associated to physiological constancy preservation, variability reduction and mostly adaptability.

METHODS

Whole night sleep electroencephalogram signals were recorded in 24 men: 10 untreated patients with a major depressive episode (41.70+/-8.11 years) and 14 healthy subjects (42.43+/-5.67 years). Scaling in these time series was investigated with detrended fluctuation analysis (time range: 0.16-2.00s). Scaling exponents (alpha) were determined in stage 2, slow wave sleep (stages 3 and 4) and during REM sleep. Forty-five epochs of 20s were chosen randomly in each of these stages.

RESULTS

The median values of alpha were lower in patients during stage 2 and SWS.

CONCLUSIONS

Major depressive episodes are characterized by a modification in the correlation structure of the sleep EEG time series. The finding which shows decreasing rate of the temporal correlations being different within the two groups in stage 2 and SWS provides an electrophysiologic argument that the underlying neuronal dynamics are modified during acute depression.

SIGNIFICANCE

The observed modifications in scaling behaviour in acutely depressed patients could be an explanation of the sleep fragmentation and instability found during major depressive episode.