Article

EEG complexity as a measure of depth of anesthesia for patients

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

A new approach for quantifying the relationship between brain activity patterns and depth of anesthesia (DOA) is presented by analyzing the spatio-temporal patterns in the electroencephalogram (EEG) using Lempel-Ziv complexity analysis. Twenty-seven patients undergoing vascular surgery were studied under general anesthesia with sevoflurane, isoflurane, propofol, or desflurane. The EEG was recorded continuously during the procedure and patients' anesthesia states were assessed according to the responsiveness component of the observer's assessment of alertness/sedation (OAA/S) score. An OAA/S score of zero or one was considered asleep and two or greater was considered awake. Complexity of the EEG was quantitatively estimated by the measure C(n), whose performance in discriminating awake and asleep states was analyzed by statistics for different anesthetic techniques and different patient populations. Compared with other measures, such as approximate entropy, spectral entropy, and median frequency, C(n) not only demonstrates better performance (93% accuracy) across all of the patients, but also is an easier algorithm to implement for real-time use. The study shows that C(n) is a very useful and promising EEG-derived parameter for characterizing the (DOA) under clinical situations.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... This transformation implicitly assumes that the dynamics of neural activity are stationary and linear, but do not consider the non-stationary and non-linear or chaotic behaviors in EEG signals [13]. To overcome these problems, many nonlinear analysis methods are proposed, for example, Lempel-Ziv complexity [14], detrended fluctuation analysis [15][16][17], fractal-scaling analysis [18], the Hurst exponent [19] and Poincaré plot [20], etc. In particular, entropies are another proposed nonlinear methods to quantify the regularity of EEG for estimating DOA, such as approximate entropy (ApEn) [21], sample entropy (SampEn) [22,23] and permutation entropy (PeEn) [24][25][26]. ...
... Moreover, the selection of parameters in these nonlinear methods is a challenge task. Although the recommend values are given [14,15,24], optimal parameters for each measure may be not suitable for all EEG series due to interpatient variability. To sum up, many features in diverse domains have been proposed for DOA assessment over past years, but the reliability of these monitors has been questioned in some special cases [29,30]. ...
... The colours clearly show the alterations in EEG spectral power. The beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) rhythms decrease with the loss of consciousness, in contrast, the slow oscillations including alpha (8)(9)(10)(11)(12) and delta (<4 Hz) activity increase. All EEGV indexes can track the changes in consciousness level with increasing anesthetic drug effect as shown in figure 10(b). ...
Article
In this paper, a new approach of extracting and measuring the variability in electroencephalogram (EEG) was proposed to assess the depth of anesthesia (DOA) under general anesthesia. The EEG variability (EEGV) was extracted as a fluctuation in time interval that occurs between two local maxima of EEG. Eight parameters related to EEGV were measured in time and frequency domains, and compared with state-of-the-art DOA estimation parameters, including sample entropy, permutation entropy, median frequency and spectral edge frequency of EEG. The area under the receiver-operator characteristics curve (AUC) and Pearson correlation coefficient were used to validate its performance on 56 patients. Our proposed EEGV-derived parameters yield significant difference for discriminating between awake and anesthesia stages at a significance level of 0.05, as well as improvement in AUC and correlation coefficient on average, which surpasses the conventional features of EEG in detection accuracy of unconscious state and tracking the level of consciousness. To sum up, EEGV analysis provides a new perspective in quantifying EEG and corresponding parameters are powerful and promising for monitoring DOA under clinical situations.
... Since, as will be made clear, entropy is a reflection of complexity, the explanation of the brain activities' entropy starts with an outline of the brains' complex activity. Zhang, Roy and Jensen (2001) state that EEG data exhibits strong non-linear and dynamical ...
... States between awake and asleep have an intermediate complexity value with gradual scaling". (Zhang et al., 2001(Zhang et al., , p. 1431 This dynamical complexity is calculable with complexity analyses (Zhang et al., 2001) and assignable by complexity measurement parameters like entropy. Hence, the entropy measurements in question assign the complexity of the brain's activity. ...
... States between awake and asleep have an intermediate complexity value with gradual scaling". (Zhang et al., 2001(Zhang et al., , p. 1431 This dynamical complexity is calculable with complexity analyses (Zhang et al., 2001) and assignable by complexity measurement parameters like entropy. Hence, the entropy measurements in question assign the complexity of the brain's activity. ...
Thesis
Full-text available
Master's Thesis by Tomko Settgast, supervised by Mathias Hegele and Dominik Endres: The text at hand investigates the possibilities to capture the so-called flow-experience without the reliance on subjective reports, i.e. it follows the intention to explore reliably and objectively measurable markers. This search especially regards neural correlates of the experience in question that have not been found yet. The text is thereby subsumed under the umbrella of enactivism since it gives similar credit to phenomenology and neuroscience, uses the description of the dynamical system’s theory and bridges the phenomenal and natural scientific aspects of cognition via an ecologically psychological sense-making. The introduction of the neurophenomenological method at the beginning of the text offers the possibility to suggest objective markers of subjective experience based on correlation. A central position within thisobjectification is given to the entropic brain hypothesis, as it is prominently represented by Carhart-Harris (2018). Its claim to connect subjective experience to the brain’s dynamically working mechanism enables its linkage to a dynamical system’s account for cognition. The dynamical attractors that the systems theory suggest for guiding behaviour is thereby easily integrated within the notions of predictive coding (i.a. Kilner, Friston, Frith, 2007; Clark, 2015) that assumes predictions to be the foundation of perception. Under the assumption of enactivism and its notion of a unity of perception and action, one gets the opportunity to translate the dynamical system’s attractors with Gibson’s (1986) idea of affordances. Thus, the brain’s dynamical working mechanism is the reflection of the phenomenal experience of affordances that guide perception and action. Especially, skilled action will be explored as the consequence of simultaneously attracting affordances which allow for the use of different strategies in pursuing a goal. This dynamically metastable attunement to different affordances (Bruineberg & Rietveld, 2014) constitutes exactly the entirety of the introduced dynamical attractors and is reflected in the brain activities’ entropy. This hypothesis is completed with the introduction of the serotonin’s and dopamine’s neuromodulation on these attractor-based affordances where those neuromodulator’s influences in perceptual guidance and behavioural selection as well as execution are emphasised. The exploration of these neurophysiological measurements enables the linkage of subjective and objective markers of flow-experience after a flow-experience’s phenomenal characterisation is given. Therefore, the outlined objective measurements are followed by an introduction of flow-experience within the notion of Csíkszentmihályi (1975). Its phenomenal characterisation and the introduced theories are used to suggest an objective measurement of flow-experience. The text uses the similarity between the flow-experience’sphenomenology and the experience of (musical) improvisation to infer a way to investigate ojectively measurable markers of flow. As it will be revealed later, this is based on the fact that the cognitive neuroscience of improvisation leads to a phenomenological experience that is summarised as the creator-witness phenomenon a fter Berkowitz (2010) what will be made fruitful as a way to investigate the state of flow. Taken together, professional musical improvisation as a specific example of skilled action shows a phenomenal proximity to flow-experience wherefore its underlying neural mechanism isinferred as the underlying mechanism of flow-experience. Hence, an objectively measurable marker of theflow’s state of mind will be explored in the increase of the brain activities’ entropy that reflects an increase in themetastable attunement to different but simultaneously visible affordances.
... In this context, the Lempel-Ziv complexity (LZC) may overcome such limitations. Indeed, LZC is capable of detecting complexity even if the time series is generated from a chaotic system or from a stochastic process, as well as from a mixture of them (i.e., the model independence property [28]). Furthermore, LZC was found to be robust to non-stationary time-series and short data [29]. ...
... Specifically, we acquired EEG signals of HS, MS, and LS, under both CE/OE rest and CE/OE hypnosis. We analyzed EEG Power Spectral Density (PSD), Lempel-Ziv Complexity (LZC), and frequency-banded Lempel-Ziv Complexity (fLZC) for 19 electrodes sites in four frequency bands: δ (1-4 Hz), θ (4-8 Hz), α (8)(9)(10)(11)(12), and β (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30). Then, for each group of subjects and each eye condition, we compared these measures between rest and hypnosis. ...
... where b(n) = n/log 2 (n) corrects for the bias on c(n) introduced by the length of the sequence [28,58]. Previous applications of LZC reported less complex and more regular EEG time series in the case of patients affected by Alzheimer's disease, as compared to control subjects [59], and in the case of anaesthetized patients, as compared to the awake state [28]. ...
Article
Full-text available
Hypnotic susceptibility is a major factor influencing the study of the neural correlates of hypnosis using EEG. In this context, while its effects on the response to hypnotic suggestions are undisputed, less attention has been paid to “neutral hypnosis” (i.e., the hypnotic condition in absence of suggestions). Furthermore, although an influence of opened and closed eye condition onto hypnotizability has been reported, a systematic investigation is still missing. Here, we analyzed EEG signals from 34 healthy subjects with low (LS), medium (MS), and (HS) hypnotic susceptibility using power spectral measures (i.e., TPSD, PSD) and Lempel-Ziv-Complexity (i.e., LZC, fLZC). Indeed, LZC was found to be more suitable than other complexity measures for EEG analysis, while it has been never used in the study of hypnosis. Accordingly, for each measure, we investigated within-group differences between rest and neutral hypnosis, and between opened-eye/closed-eye conditions under both rest and neutral hypnosis. Then, we evaluated between-group differences for each experimental condition. We observed that, while power estimates did not reveal notable differences between groups, LZC and fLZC were able to distinguish between HS, MS, and LS. In particular, we found a left frontal difference between HS and LS during closed-eye rest. Moreover, we observed a symmetric pattern distinguishing HS and LS during closed-eye hypnosis. Our results suggest that LZC is better capable of discriminating subjects with different hypnotic susceptibility, as compared to standard power analysis.
... In particular, the use of EEG for understanding the brain's complex dynamics has been growing in popularity. Indeed, there has been an acceleration in the number of studies attempting to delineate maladaptive psychopathological mechanisms by investigating EEG complexity in various experimental conditions, from sleeping (e.g., Chouvarda et al., 2010) and anaesthetic states (e.g., Zhang et al., 2001), to tasks that involve responding to emotional stimuli (e.g., Aftanas, Lotova, Koshkarov, Popov, et al., 1997) and mental arithmetic (e.g., M. R. Mohammadi et al., 2016). Complexity measures of EEG signals may supplement, if not provide greater utility and sensitivity than conventional EEG analysis techniques (such as event-related potentials or time/frequency analysis) in detecting changes in psychopathological states (Sohn et al., 2010) and potentially expediting diagnosis of diseases (Czigler et al., 2008). ...
... Simply put, the regularity of the signal is determined through scanning the symbolic sequences for new patterns, increasing the complexity count every time a new sequence is detected. As compared to other complexity measures, LZC is relatively less computationally expensive, easy to implement, and can be applied directly on biological signals without any preprocessing steps (Aiordachioaie & Popescu, 2020; Zhang et al., 2001). Therefore, even though the performance of LZC, as relative to other indices, remains indeterminate (Aiordachioaie & Popescu, 2020;Fathillah et al., 2017;Ibanez-Molina et al., 2015), it has been applied extensively in various fields. ...
... Apart from sleep stages (for a comprehensive review, see Ma et al., 2018), there is also convincing evidence showing reduced EEG complexity in other states of relative "loss" of consciousness (see Table 4), including anesthetized states (M. Schartner et al., 2015;Zhang et al., 2001), seizures (Kannathal et al., 2005;Krystal et al., 1996) and disorders of consciousness (i.e., vegetative and minimally conscious states, see Perturbational Complexity Index -PCI, Casali et al., 2013). ...
Preprint
Full-text available
While electroencephalography (EEG) signals are commonly examined using conventional linear methods, there has been an increasing trend towards the use of complexity analysis in quantifying neural activity. On top of revealing complex neuronal processes of the brain that may not be possible with linear approaches, EEG complexity measures have also demonstrated their potential as biomarkers of psychopathology such as depression and schizophrenia. Unfortunately, the opacity of algorithms and descriptions originating from mathematical concepts have made it difficult to understand what complexity is and how to draw consistent conclusions when applied within psychology and neuropsychiatry research. In this review, we provide an overview and entry-level explanation of existing EEG complexity measures, which can be broadly categorized as measures of (1) predictability and (2) regularity. We then synthesize complexity findings across different areas of psychological science, namely in consciousness research, mood and anxiety disorders, schizophrenia, neurodevelopmental and neurodegenerative disorders, as well as changes across the lifespan, while addressing some theoretical and methodological issues underlying the discrepancies in the data. Finally, we present important considerations when choosing and interpreting these metrics.
... During the transition from wakefulness to sleep or anesthesia, the number of possible experiences and cognitive processes that one can have is greatly reduced, and thus it is coherent that the complexity of brain activity follows the same pattern. In fact, this reduction of the repertoire of brain activity has been seen in rats Although 1/f slope and LZc have distant mathematical origins, one coming from spectral analysis and the other one from Information Theory, both have been shown to correlate with GSC (Miskovic et al., 2019;Zhang et al., 2001). We hypothesize that this could be due to an underlying intrinsic relation between E/I balance and the repertoire of activity patterns in cortical systems. ...
... On the other hand, applying a ff corresponds to setting the amplitude of every frequency higher than ff to zero. To maintain time series stationarity, a requirement of the LZc algorithm Zhang et al., 2001), all iDFT models were made with a f0 = 1Hz unless otherwise stated. For every set of simulations, we generated a 256 time series with different values of s. ...
... ing sleep and during propofol(Zhang et al., 2001; or xenon(Sarasso et al., 2015) anaesthesia. This has been stated as support for the Entropic Brain Hypothesis (Carhart-Harris et al., 2014) , which proposes that the complexity (entropy) of brain activity should directly reflect the complexity of states of consciousness (experience). ...
Thesis
Full-text available
Spontaneous fluctuations occur at different spatial and temporal scales in the brain. Depending on its scale, these activities can show characteristic hallmarks. From a mesoscale perspective, in spontaneous conditions, cortical neurons fire action potentials in a seemingly stochastic manner, which extrapolated to an entire population shows a dynamical state coined as the asynchronous irregular state. Interestingly, when a local population of balanced excitation and inhibition is recurrently connected, the synchronous population generates a baseline of stochastic perturbation over the neuron’s membrane potential of that local population. These perturbations have been proposed as optimal for information computation and are associated with different states of attention at the behavioral level. Specifically, Locus-Coeruleus Noradrenergic (LC-NE) neuromodulation -which regulates brain states- has been highly implicated in the modulation of desynchronized activity. In this dissertation, we will use a modeling-driven analysis of attentional modulation of local electrophysiological desynchronization, hypothesizing that LC-NE neuromodulation shapes desynchronized background state and the balance between excitation and inhibition. We will show how the complexity of the electrophysiological signals depends on the excitation-inhibition balance of cortical activity, how it tracks behavioral performance, and how it can be related to LC-NE activity and arousal-related neuromodulation. Finally, we show how this complexity fluctuates at different spatial scales with low-dimensionality in attention, and how it is tracked by pupil diameter fluctuations -a non-invasive proxy of LC-NE activity and arousal- in a visuospatial working memory task in humans.
... 17,19 In general, complexity measures are higher during the awake state, in which the brain is more active than in the asleep state. 20 Notably, less complexity and irregularity of brain rhythms are observed in comatose patients. 20,21 Thus, the EEG complexity measures of a particular brain state, such as the resting state, can be used to evaluate brain activity in comatose patients. ...
... 20 Notably, less complexity and irregularity of brain rhythms are observed in comatose patients. 20,21 Thus, the EEG complexity measures of a particular brain state, such as the resting state, can be used to evaluate brain activity in comatose patients. Different brain networks are engaged in consciousness and information processing during wakefulness. ...
... Complexity feature analysis can be applied to differentiate the states of the brain. 20,21 Previous studies also observed less complexity and irregularity of brain wave activity in comatose patients. 20,21,81 Accordingly, our findings support this evidence and our hypothesis that nasal AP can change brain signals into a more activated state. ...
Article
Objectives Coma state and loss of consciousness are associated with impaired brain activity, particularly gamma oscillations, that integrate functional connectivity in neural networks, including the default mode network (DMN). Mechanical ventilation (MV) in comatose patients can aggravate brain activity, which has decreased in coma, presumably because of diminished nasal airflow. Nasal airflow, known to drive functional neural oscillations, synchronizing distant brain networks activity, is eliminated by tracheal intubation and MV. Hence, we proposed that rhythmic nasal air puffing in mechanically ventilated comatose patients may promote brain activity and improve network connectivity. Materials and Methods We recorded electroencephalography (EEG) from 15 comatose patients (seven women) admitted to the intensive care unit because of opium poisoning and assessed the activity, complexity, and connectivity of the DMN before and during the nasal air-puff stimulation. Nasal cavity air puffing was done through a nasal cannula controlled by an electrical valve (open duration of 630 ms) with a frequency of 0.2 Hz (ie, 12 puff/min). Results Our analyses demonstrated that nasal air puffing enhanced the power of gamma oscillations (30–100 Hz) in the DMN. In addition, we found that the coherence and synchrony between DMN regions were increased during nasal air puffing. Recurrence quantification and fractal dimension analyses revealed that EEG global complexity and irregularity, typically seen in wakefulness and conscious state, increased during rhythmic nasal air puffing. Conclusions Rhythmic nasal air puffing, as a noninvasive brain stimulation method, opens a new window to modifying the brain connectivity integration in comatose patients. This approach may potentially influence comatose patients’ outcomes by increasing brain reactivity and network connectivity.
... Within the particular field of anaesthetics research, measures such as Permutation Entropy [9] have been used to quantify the effect of sevoflurane on EEG signals, obtaining better results than classical measures such as the Bispectral Index (BIS) [10][11][12]. Lempel-Ziv complexity has been used to study EEG signals in patients under the effects of propofol [13]. Another study compares different entropic measures in patients under anaesthesia induced by GABAergic agents [14]. ...
... The hypothesis in which the dynamic state of higher-than-normal entropy might correspond to a psychedelic or hallucinatory state of consciousness has become known as the Entropic Brain Hypothesis [48,49]. Moreover, these results are consistent with other studies showing that Lempel-Ziv complexity and entropy increase under effect of ketamine [13,47,50]. ...
Preprint
Full-text available
The use of anaesthesia is a fundamental tool in the investigation of consciousness. Anesthesia procedures allow to investigate different states of consciousness from sedation to deep anesthesia within controlled scenarios. In this study we use information quantifiers to measure the complexity of electrocorticogram recordings in monkeys. We apply these metrics to compare different stages of general anesthesia for evaluating consciousness in several anesthesia protocols. We find that the complexity of brain activity can be used as a correlate of consciousness. For two of the anaesthetics used, propofol and medetomidine, we find that the anaesthetised state is accompanied by a reduction in the complexity of brain activity. On the other hand we observe that use of ketamine produces an increase in complexity measurements. We relate this observation with increase activity within certain brain regions associated with the ketamine used doses. Our measurements indicate that complexity of brain activity is a good indicator for a general evaluation of different levels of consciousness awareness, both in anesthetized and non anesthetizes states.
... In particular, LempeleZiv Complexity, which can be used to compute the number of non-redundant patterns of EEG signals, 13 has shown particularly promising results in discriminating global states of consciousness. The pioneer work of Zhang and colleagues 14 showed that, using a simple threshold value, they could discriminate between awake (higher) and deep anaesthesia (lower) LempeleZiv complexity using four different anaesthetics with 93% accuracy. More recent work has also found that reverberations of brain activity evoked by transcranial magnetic stimulation are much more complex during normal wakefulness than during sleep, or after loss of consciousness attributable to a variety of anaesthetic drugs. ...
... Previous research has shown that LempeleZiv complexity (or similar measures) decreases after loss of consciousness during sleep 16 and during propofol 14,23 or xenon 16 anaesthesia. This has been used as support for the entropic brain hypothesis, 26 which proposes that the complexity (entropy) of brain activity should directly reflect the complexity of states of consciousness (experience). ...
Article
Background Brain activity complexity is a promising correlate of states of consciousness. Previous studies have shown higher complexity for awake compared with deep anaesthesia states. However, little attention has been paid to complexity in intermediate states of sedation. Methods We analysed the Lempel–Ziv complexity of EEG signals from subjects undergoing moderate propofol sedation, from an open access database, and related it to behavioural performance as a continuous marker of the level of sedation and to plasma propofol concentrations. We explored its relation to spectral properties, to propofol susceptibility, and its topographical distribution. Results Subjects who retained behavioural performance despite propofol sedation showed increased brain activity complexity compared with baseline (M=13.9%, 95% confidence interval=7.5–20.3). This was not the case for subjects who lost behavioural performance. The increase was most prominent in frontal electrodes, and correlated with behavioural performance and propofol susceptibility. This effect was positively correlated with high-frequency activity. However, abolishing specific frequency ranges (e.g. alpha or gamma) did not reduce the propofol-induced increase in Lempel–Ziv complexity. Conclusions Brain activity complexity can increase in response to propofol, particularly during low-dose sedation. Propofol-mediated Lempel–Ziv complexity increase was independent of frequency-specific spectral power manipulations, and most prominent in frontal areas. Taken together, these results advance our understanding of brain activity complexity and anaesthetics. They do not support models of consciousness that propose a direct relation between brain activity complexity and states of consciousness.
... Hutt [6] has deployed a linear neural population model which predicts the concentration of anesthetic propofol using the power spectrum of EEG signals. Zhang et al. [7] have adopted spatio-temporal patterns in the electroencephalogram (EEG) using Lempel-Ziv analysis. Various pattern recognition methods for different cognitive task classification were carried out in [8] with an accuracy of 93% using machine learning algorithms. ...
... With the already research done in [15], it is evident that there is a correlation between the brain wave activity at different frequency components of the EEG signals and different phenomena. From a clinical point of view, the raw EEG has usually been described in terms of frequency bands: gamma (greater than 30 Hz), beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), alpha (8)(9)(10)(11)(12), theta (4)(5)(6)(7)(8), and delta (less than 4 Hz). With the induction of anesthetics there is a significant drop in the activity of the high frequency beta and alpha bands while there is an increased response observed in low frequency band during deep anesthesia level [16]. ...
Article
According to a recently conducted survey on surgical complication mortality rate, 47% of such cases are due to anesthetics overdose. This indicates that there is an urgent need to moderate the level of anesthesia. Recently deep learning (DL) methods have played a major role in estimating the depth of Anesthesia (DOA) of patients and has played an essential role in control anesthesia overdose. In this paper, Electroencephalography (EEG) signals have been used for the prediction of DOA. EEG signals are very complex signals which may require months of training and advanced signal processing techniques. It is a point of debate whether DL methods are an improvement over the already existing traditional EEG signal processing approaches. One of the DL algorithms is Convolutional neural network (CNN) which is very popular algorithm for object recognition and is widely growing its applications in processing hierarchy in the human visual system. In this paper, various decomposition methods have been used for extracting the features EEG signal. After acquiring the necessary signals values in image format, several CNN models have been deployed for classification of DOA depending upon their Bispectral Index (BIS) and the signal quality index (SQI). The EEG signals were converted into the frequency domain using and Empirical Mode Decomposition (EMD), and Ensemble Empirical Mode Decomposition (EEMD). However, because of the inter mode mixing observed in EMD method; EEMD have been utilized for this study. The developed CNN models were used to predict the DOA based on the EEG spectrum images without the use of handcrafted features which provides intuitive mapping with high efficiency and reliability. The best trained model gives an accuracy of 83.2%. Hence, this provides further scope and research which can be carried out in the domain of visual mapping of DOA using EEG signals and DL methods.
... Within the particular field of anaesthetics research, measures such as Permutation Entropy Keller et al. (2017) have been used to quantify the effect of sevoflurane on EEG signals, obtaining better results than classical measures such as the Bispectral Index (BIS) Todd (1998), Li et al. (2008Li et al. ( , 2010. Lempel-Ziv complexity has been used to study EEG signals in patients under the effects of propofol Zhang et al. (2001). Another study compares different entropic measures in patients under anaesthesia induced by GABAergic agents Liang et al. (2015). ...
... The hypothesis in which the dynamic state of higher-than-normal entropy might correspond to a psychedelic or hallucinatory state of consciousness has become known as the Entropic Brain Hypothesis Carhart-Harris and Friston (2019), Carhart-Harris et al. (2014). Moreover, these results are consistent with other studies showing that Lempel-Ziv complexity and entropy increase under effect of ketamine Schartner et al. (2017), Liu et al. (2018), Zhang et al. (2001. In other electrophysiological studies, ketamine affects sensory gating and alters the oscillatory characteristics of neuronal signals in a complex manner Lazarewicz et al. (2010). ...
Article
Full-text available
The use of anaesthesia is a fundamental tool in the investigation of consciousness. Anesthesia procedures allow to investigate different states of consciousness from sedation to deep anesthesia within controlled scenarios. In this study we use information quantifiers to measure the complexity of electrocorticogram recordings in monkeys. We apply these metrics to compare different stages of general anesthesia for evaluating consciousness in several anesthesia protocols. We find that the complexity of brain activity can be used as a correlate of consciousness. For two of the anaesthetics used, propofol and medetomidine, we find that the anaesthetised state is accompanied by a reduction in the complexity of brain activity. On the other hand we observe that use of ketamine produces an increase in complexity measurements. We relate this observation with increase activity within certain brain regions associated with the ketamine used doses. Our measurements indicate that complexity of brain activity is a good indicator for a general evaluation of different levels of consciousness awareness, both in anesthetized and non anesthetizes states.
... A total of 10 cross-validations were implemented for the evaluation of the models [38]. Here, the data is divided into 10 equally distributed segments, where nine of them are used for training and the remaining one is used for testing. ...
... Unsupervised learning reduces the data pre-processing (i.e., labeling the image), where many skilled quality engineers have traditionally been dedicated to this work. The introduction of anomaly detection is a game changing method that is more powerful and faster than other supervised learning methods [38][39][40]. A pre-trained ResNet-18 was used in this research to accelerate the inference speed, because it is faster and more compact than other backbone algorithms, such as ResNet34, ResNet50, ResNet101, and ResNet152. ...
Article
Full-text available
Deep learning methods are currently used in industries to improve the efficiency and quality of the product. Detecting defects on printed circuit boards (PCBs) is a challenging task and is usually solved by automated visual inspection, automated optical inspection, manual inspection, and supervised learning methods, such as you only look once (YOLO) of tiny YOLO, YOLOv2, YOLOv3, YOLOv4, and YOLOv5. Previously described methods for defect detection in PCBs require large numbers of labeled images, which is computationally expensive in training and requires a great deal of human effort to label the data. This paper introduces a new unsupervised learning method for the detection of defects in PCB using student–teacher feature pyramid matching as a pre-trained image classification model used to learn the distribution of images without anomalies. Hence, we extracted the knowledge into a student network which had same architecture as the teacher network. This one-step transfer retains key clues as much as possible. In addition, we incorporated a multi-scale feature matching strategy into the framework. A mixture of multi-level knowledge from the features pyramid passes through a better supervision, known as hierarchical feature alignment, which allows the student network to receive it, thereby allowing for the detection of various sizes of anomalies. A scoring function reflects the probability of the occurrence of anomalies. This framework helped us to achieve accurate anomaly detection. Apart from accuracy, its inference speed also reached around 100 frames per second.
... There is increasing evidence for a strong association between neural information measures, such as electrophysiological signal complexity, and level of consciousness (Ab asolo et al., 2015;Castro-Zaballa et al., 2019;Gonz alez et al., 2019;Mateos et al., 2018;Schartner et al., 2015;Schartner, Carhart-Harris, et al., 2017;Zhang et al., 2001). One of the most studied neural complexity metrics is Lempel-Ziv (LZ) complexity, capturing the number of distinct substrings or patterns within a sequence (Lempel & Ziv, 1976;Ziv & Lempel, 1978). ...
... One of the most studied neural complexity metrics is Lempel-Ziv (LZ) complexity, capturing the number of distinct substrings or patterns within a sequence (Lempel & Ziv, 1976;Ziv & Lempel, 1978). A decrease in complexity has been demonstrated for anaesthesia (Li & Mashour, 2019;Schartner et al., 2015;Zhang et al., 2001) and during non-rapid eye movement sleep (NREM sleep) when compared with normal wakefulness. However, REM complexity has consistently been shown to be above NREM sleep and below normal wakefulness (Ab asolo et al., 2015;Andrillon et al., 2016;Mateos et al., 2018;Schartner, Pigorini, et al., 2017). ...
Article
Full-text available
There is increasing evidence that level of consciousness can be captured by neural informational complexity: for instance, complexity, as measured by the Lempel Ziv (LZ) compression algorithm, decreases during anesthesia and non‐rapid eye movement (NREM) sleep in humans and rats, when compared to LZ in awake and REM sleep. In contrast, LZ is higher in humans under the effect of psychedelics, including subanesthetic doses of ketamine. However, it is both unclear how this result would be modulated by varying ketamine doses, and whether it would extend to other species. Here we studied LZ with and without auditory stimulation during wakefulness and different sleep stages in 5 cats implanted with intracranial electrodes, as well as under subanesthetic doses of ketamine (5, 10, and 15 mg/kg i.m.). In line with previous results, LZ was lowest in NREM sleep, but similar in REM and wakefulness. Furthermore, we found an inverted U‐shaped curve following different levels of ketamine doses in a subset of electrodes, primarily in prefrontal cortex. However, it is worth noting that the variability in the ketamine dose‐response curve across cats and cortices was larger than that in the sleep‐stage data, highlighting the differential local dynamics created by two different ways of modulating conscious state. These results replicate previous findings, both in humans and other species, demonstrating that neural complexity is highly sensitive to capture state changes between wake and sleep stages while adding a local cortical description. Finally, this study describes the differential effects of ketamine doses, replicating a rise in complexity for low doses, and further fall as doses approach anesthetic levels in a differential manner depending on the cortex.
... The nonlinear Lempel-Ziv complexity (LZC), introduced by Lempel and Ziv [26], measures the complexity of a signal and has been successfully used on EEG signals for the detection of different mental states [27,28]. EEG data from severe Alzheimer's disease patients showed a loss of complexity over a wide range of time scales, indicating a destruction of nonlinear structures in brain dynamics [29][30][31]. ...
... One of the most widely used methods to analyse EEG signals is to decompose the signal into functionally distinct frequency bands, such as delta (1-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30), and gamma (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45). In the current study, this was achieved by first calculating the power spectral density of the EEG signal by Welch's method, as done by Bachmann et al. 2018. ...
Article
Full-text available
Depression is a public health issue that severely affects one’s well being and can cause negative social and economic effects to society. To raise awareness of these problems, this research aims at determining whether the long-lasting effects of depression can be determined from electroencephalographic (EEG) signals. The article contains an accuracy comparison for SVM, LDA, NB, kNN, and D3 binary classifiers, which were trained using linear (relative band power, alpha power variability, spectral asymmetry index) and nonlinear (Higuchi fractal dimension, Lempel–Ziv complexity, detrended fluctuation analysis) EEG features. The age- and gender-matched dataset consisted of 10 healthy subjects and 10 subjects diagnosed with depression at some point in their lifetime. Most of the proposed feature selection and classifier combinations achieved accuracy in the range of 80% to 95%, and all the models were evaluated using a 10-fold cross-validation. The results showed that the motioned EEG features used in classifying ongoing depression also work for classifying the long-lasting effects of depression.
... Remarkably, during wakefulness, feedback is higher than feedforward connectivity, and this feedback dominance is significantly reduced during anesthesia (Lee et al., 2009). The complexity of the signal of the EEG can also be useful to predict the depth of anesthesia (Zhang et al., 2001). Studies using Lempel Ziv Complexity analysis (and other algorithms to study brain entropy) revealed that general anesthesia decreases EEG signal complexity (Liang et al., 2015;Hudetz et al., 2016). ...
Chapter
On a daily basis, our brain alternates between several states of vigilance (or arousal) that are internally generated or, more often, generated in response to environmental cues and challenges. In this chapter, we focus on states of sleep and wakefulness, general anesthesia, as well as other nonpathological states of vigilance that occur in response to extreme environmental conditions such as torpor and hibernation. Each of these states has several unique and distinctive parameters that can be objectively assessed by observation (body posture and behaviors), as well as with more sophisticated analytical approaches (cardiorespiratory and electroencephalographic features). Here we provide operational definitions and distinctive characteristics, and introduce some quantitative analytical tools used in research and clinical settings to study these states of vigilance.
... In this regard, have shown that complexity, assessed by permutation entropy, is higher during W than both sleep states . Furthermore, EEG complexity decreases under general anesthesia (Zhang et al., 2001;Hudetz et al., 2016). ...
Preprint
Full-text available
Urethane is a general anesthetic widely used in animal research. It is unique among anesthetics because urethane anesthesia alternates between macroscopically distinct electrographic states: a slow-wave state that resembles NREM sleep (NREMure), and an activated state with features of both REM sleep and wakefulness (REMure). However, the relationship between urethane anesthesia and physiological sleep is still unclear. In this study, electroencephalography (EEG) and electromyography were recorded in chronically prepared rats during natural sleep-wake states and during urethane anesthesia. We subsequently analyzed the EEG signatures associated with the loss of consciousness and found that, in comparison to natural sleep-wake states, the power, coherence, directed connectivity and complexity of brain oscillations are distinct during urethane. We also demonstrate that both urethane states have clear EEG signatures of general anesthesia. Thus, despite superficial similarities that have led others to conclude that urethane is a model of sleep, the electrocortical traits of depressed and activated states during urethane anesthesia differ from physiological sleep states.
... We use the Cambridge Centre for Ageing and Neuroscience (CAMCAN) dataset [12], which includes a large-scale MEG dataset of participants undergoing several cognitive tasks, and study the differences in Lempel-Ziv complexity [1] between participants in wakeful rest, and participants performing a simple cognitive stop-signal, go/no-go task [12]. This measure (or minor variations of it) has been widely used in the neuroscience literature [13][14][15][16], showing a remarkable performance in discriminating between different states of consciousness, for instance normal wakefulness versus sleep [3]. ...
Article
Full-text available
When employing nonlinear methods to characterize complex systems, it is important to determine to what extent they are capturing genuine nonlinear phenomena that could not be assessed by simpler spectral methods. Specifically, we are concerned with the problem of quantifying spectral and phasic effects on an observed difference in a nonlinear feature between two systems (or two states of the same system). Here we derive, from a sequence of null models, a decomposition of the difference in an observable into spectral, phasic, and spectrum-phase interaction components. Our approach makes no assumptions about the structure of the data and adds nuance to a wide range of time series analyses.
... Data sample above the threshold is equated to 1 and below the threshold level to 0. The resulting binary segment is scanned for different patterns. The counter c(n) is increased by one unit when a new pattern is encountered in the scanning process (Zhang et al., 2001). ...
Article
Full-text available
Background Functional connectivity and complexity analysis has been discretely studied to understand intricate brain dynamics. The current study investigates the interplay between functional connectivity and complexity using the Kuramoto mean-field model. Method Functional connectivity matrices are estimated using the weighted phase lag index and complexity measures through popularly used complexity estimators such as Lempel-Ziv complexity (LZC), Higuchi's fractal dimension (HFD), and fluctuation-based dispersion entropy (FDispEn). Complexity measures are estimated on real and simulated electroencephalogram (EEG) signals of patients with mild cognitive-impaired Alzheimer's disease (MCI-AD) and controls. Complexity measures are further applied to simulated signals generated from lesion-induced connectivity matrix and studied its impact. It is a novel attempt to study the relation between functional connectivity and complexity using a neurocomputational model. Results Real EEG signals from patients with MCI-AD exhibited reduced functional connectivity and complexity in anterior and central regions. A simulation study has also displayed significantly reduced regional complexity in the patient group with respect to control. A similar reduction in complexity was further evident in simulation studies with lesion-induced control groups compared with non-lesion-induced control groups. Conclusion Taken together, simulation studies demonstrate a positive influence of reduced connectivity in the model imparting a reduced complexity in the EEG signal. The study revealed the presence of a direct relation between functional connectivity and complexity with reduced connectivity, yielding a decreased EEG complexity.
... In a previous report, the complexity of EEG estimated for assess patients' anesthesia under sevoflurane, isoflurane, propofol, or desflurane, pursuant to the behavior of observer's assessment of alertness/sedation (OAA/S) score. LZC index compared with other nonparametric measures demonstrated precision of more than 90% in estimating action from brain function [70]. Indeed, we found that VNS can decrease the amplitude of slow waves in S1 significantly, but the changes in LFP slow waves of V1 were not considerable. ...
Article
Recent studies suggest that vagus nerve stimulation (VNS) promotes cognitive and behavioral restoration after traumatic brain injuries. As vagus nerve has wide effects over the brain and visceral organs, stimulation of the sensory/visceral afferents might have a therapeutic potential to modulate the level of consciousness. One of the most important challenges in studying consciousness is objective evaluation of the consciousness level. Brain complexity that can be measured through Lempel-Ziv complexity (LZC) index was used as a novel mathematical approach for objective measurement of consciousness. The main goal of our study was to examine the effects of VNS on LZC index of consciousness. In this study, we did VNS on the anesthetized rats, and simultaneously LFPs recording was performed in two different cortical areas of primary somatosensory (S1) or visual (V1) cortex. LZC and the amplitude of slow waves were computed during different periods of VNS. We found that the LZC index during VNS period was significantly higher in both of the cortical areas of S1 and V1. Slow-wave activity decreased during VNS in S1, while there was no significant change in V1. Our findings showed that VNS can augment the consciousness level, and LZC index is a more sensitive parameter for detecting the level of consciousness.
... This result further highlights a decrease in complexity brought forward by deep anesthesia. Few spatial filters (i.e., 5 microstates) can better explain the ongoing EEG, in line with results suggesting a reduction in complexity to be a predictor of unconsciousness (Dasilva et al., 2020;Zhang et al., 2001). Considering that a winner-takes-all strategy was used to estimate microstate topographies (see Methods), an increased duration combined with increased spatial correlation is suggestive of (which was not certified by peer review) is the author/funder. ...
Preprint
Full-text available
It is commonly believed that the stream of consciousness is not continuous but parsed into transient brain states manifesting themselves as discrete spatiotemporal patterns of global neuronal activity. Electroencephalographical (EEG) microstates are proposed as the neurophysiological correlates of these transiently stable brain states that last for fractions of seconds. To further understand the link between EEG microstate dynamics and consciousness, we continuously recorded high-density EEG in 23 surgical patients from their awake state to unconsciousness, induced by step-wise increasing concentrations of the intravenous anesthetic propofol. Besides the conventional parameters of microstate dynamics, we introduce a new method that estimates the complexity of microstate sequences. The brain activity under the surgical anesthesia showed a decreased sequence complexity of the stereotypical microstates, which became sparser and longer-lasting. However, we observed an initial increase in microstates’ temporal dynamics and complexity with increasing depth of sedation leading to a distinctive “U-shape” that may be linked to the paradoxical excitation induced by moderate levels of propofol. Our results support the idea that the brain is in a metastable state under normal conditions, balancing between order and chaos in order to flexibly switch from one state to another. The temporal dynamics of EEG microstates indicate changes of this critical balance between stability and transition that lead to altered states of consciousness. Highlights EEG microstates capture discrete spatiotemporal patterns of global neuronal activity We studied their temporal dynamics in relation to different states of consciousness We introduce a new method to estimate the complexity of microstates sequences With moderate sedation complexity increases then decreases with full sedation Complexity of microstate sequences is sensitive to altered states of consciousness
... HFD is a fast nonlinear computational method for obtaining the fractal dimension of time series signals [96][97][98] and yields biomarkers which are significantly lower in AD patients than in normal subjects [85][86][87]. LZC is a nonparametric and nonlinear method that provides a way to quantify the complexity of the EEG [99,100] and has been used to analyse the EEG complexity in AD [101][102][103]. ...
Article
Full-text available
Biomarkers to detect Alzheimer’s disease (AD) would enable patients to gain access to appropriate services and may facilitate the development of new therapies. Given the large numbers of people affected by AD, there is a need for a low-cost, easy-to-use method to detect AD patients. Potentially, the electroencephalogram (EEG) can play a valuable role in this, but at present no single EEG biomarker is robust enough for use in practice. This study aims to provide a methodological framework for the development of robust EEG biomarkers to detect AD with clinically acceptable performance by exploiting the combined strengths of key biomarkers. A large number of existing and novel EEG biomarkers associated with slowing of EEG, reduction in EEG complexity. and decrease in EEG connectivity were investigated. Support vector machine and linear discriminate analysis methods were used to find the best combination of the EEG biomarkers to detect AD with significant performance. A total of 325,567 EEG biomarkers were investigated, and a panel of six biomarkers was identified and used to create a diagnostic model with high performance (>=85% for sensitivity and 100% for specificity).
... Loss of consciousness due to anesthesia or coma shares common features: Complexity of dynamics and neural communication are generally reduced in low-level states of consciousness (Sitt et al. 2014;Schartner et al. 2015). Consequently, estimates of complexity of human brain activity have been used to assess the depth of anesthesia (Zhang et al. 2001;Singh et al. 2017) and to predict the recovery of consciousness in vegetative patients (Sarà et al. 2011). Reduction of complexity is consistent with a deviation from critical dynamics when consciousness is lost. ...
Article
The study of states of arousal is key to understand the principles of consciousness. Yet, how different brain states emerge from the collective activity of brain regions remains unknown. Here, we studied the fMRI brain activity of monkeys during wakefulness and anesthesia-induced loss of consciousness. We showed that the coupling between each brain region and the rest of the cortex provides an efficient statistic to classify the two brain states. Based on this and other statistics, we estimated maximum entropy models to derive collective, macroscopic properties that quantify the system’s capabilities to produce work, to contain information, and to transmit it, which were all maximized in the awake state. The differences in these properties were consistent with a phase transition from critical dynamics in the awake state to supercritical dynamics in the anesthetized state. Moreover, information-theoretic measures identified those parameters that impacted the most the network dynamics. We found that changes in the state of consciousness primarily depended on changes in network couplings of insular, cingulate, and parietal cortices. Our findings suggest that the brain state transition underlying the loss of consciousness is predominantly driven by the uncoupling of specific brain regions from the rest of the network.
... The majority of these studies rely ultimately on point-summaries of 'complexity' (e.g. Lempel-Ziv complexity [17,18], entropy [22,23], etc.). However, these point-summary measures, while informative, collapse multi-scale dynamics into a single number and thus have difficulty capturing its specific shape or form. ...
Article
Full-text available
Research has found that the vividness of conscious experience is related to brain dynamics. Despite both being anaesthetics, propofol and ketamine produce different subjective states: we explore the different effects of these two anaesthetics on the structure of dynamic attractors reconstructed from electrophysiological activity recorded from cerebral cortex of two macaques. We used two methods: the first embeds the recordings in a continuous high-dimensional manifold on which we use topological data analysis to infer the presence of higher-order dynamics. The second reconstruction, an ordinal partition network embedding, allows us to create a discrete state-transition network, which is amenable to information-theoretic analysis and contains rich information about state-transition dynamics. We find that the awake condition generally had the ‘richest’ structure, visiting the most states, the presence of pronounced higher-order structures, and the least deterministic dynamics. By contrast, the propofol condition had the most dissimilar dynamics, transitioning to a more impoverished, constrained, low-structure regime. The ketamine condition, interestingly, seemed to combine aspects of both: while it was generally less complex than the awake condition, it remained well above propofol in almost all measures. These results provide deeper and more comprehensive insights than what is typically gained by using point-measures of complexity.
... This makes it useful for machine learning engineers as there exists an active and vast PyTorch community to support the researchers. YOLO-v5 is also much faster than all the previous versions of YOLO [29][30][31]. In addition to this, YOLO-v5 is nearly 90% smaller than YOLO-v4. ...
Article
In this paper, a new model known as YOLO-v5 is initiated to detect defects in PCB. In the past many models and different approaches have been implemented in the quality inspection for detection of defect in PCBs. This algorithm is specifically selected due to its efficiency, accuracy and speed. It is well known that the traditional YOLO models (YOLO, YOLO-v2, YOLO-v3, YOLO-v4 and Tiny-YOLO-v2) are the state-of-the-art in artificial intelligence industry. In electronics industry, the PCB is the core and the most basic component of any electronic product. PCB is almost used in each and every electronic product that we use in our daily life not only for commercial purposes, but also used in sensitive applications such defense and space exploration. These PCB should be inspected and quality checked to detect any kind of defects during the manufacturing process. Most of the electronic industries are focused on the quality of their product, a small error during manufacture or quality inspection of the electronic products such as PCB leads to a catastrophic end. Therefore, there is a huge revolution going on in the manufacturing industry where the object detection method like YOLO-v5 is a game changer for many industries such as electronic industries.
... It is also normalized to make it independent of the sequence length. This EEG complexity measure has already proved its ability to assess the depth of anaesthesia and it is computable in real time [57]. Hence, it can be a good feature to characterize the different sleep stages. ...
Article
Full-text available
Extensive experiments have been carried out in this study to classify sleep EEG from three different standard databases - Sleep EDF, DREAMS, and Expanded sleep EDF databases. Both two-class (sleep-awake) and multiclass classifications have been performed using a fusion of various EEG features and an ensemble classifier called random undersampling with boosting technique (RUSBoost). The results achieved using a single channel EEG are comparable to or better than the state-of-the-art methods in the literature for both types of classification, on all the databases. Two-class classification is useful to determine the preferred timings for sensory stimulation of patients with disorders of consciousness. 10-fold cross-validation accuracies of 92.6% and 97.9% have been obtained on the Sleep EDF database for 6-class and 2-class problems, respectively. Using the Expanded Sleep-EDF dataset, the accuracies improved to 96.3% for 6-state and 99.8% for 2-state classification. For the DREAMS dataset, we achieved an accuracy of 96.6% for the 2-state classification. Unlike most research in the literature where performance on unseen subjects is not considered, we report classification results on the data from unseen test subjects using both 50%-holdout and leave-one-out cross-validation approaches. Similar results were achieved using both validation techniques for different datasets emphasizing the reliability of our method. These results are very crucial for the method to be applicable for clinical use on new patients.
... It is important to note that while the primary variable returned by each function is the estimated entropy value, most functions provide secondary and tertiary variables that may be of additional interest to the user. Some [8] which also returns the reverse dispersion entropy [50], the spectral entropy function (SpecEn) [74] which also returns the band-spectral entropy [102], and the Kolmogorov entropy function (K2En) [63] which also returns the correlation sum estimate. Furthermore, every Multiscale and Multiscale Cross function has the option to plot the multiscale (cross) entropy curve (Fig 1), as well as some Base functions which allow one to plot spatial representations of the original time series (Figs 2 and 3). ...
Article
Full-text available
An increasing number of studies across many research fields from biomedical engineering to finance are employing measures of entropy to quantify the regularity, variability or randomness of time series and image data. Entropy, as it relates to information theory and dynamical systems theory, can be estimated in many ways, with newly developed methods being continuously introduced in the scientific literature. Despite the growing interest in entropic time series and image analysis, there is a shortage of validated, open-source software tools that enable researchers to apply these methods. To date, packages for performing entropy analysis are often run using graphical user interfaces, lack the necessary supporting documentation, or do not include functions for more advanced entropy methods, such as cross-entropy, multiscale cross-entropy or bidimensional entropy. In light of this, this paper introduces EntropyHub, an open-source toolkit for performing entropic time series analysis in MATLAB, Python and Julia. EntropyHub (version 0.1) provides an extensive range of more than forty functions for estimating cross-, multiscale, multiscale cross-, and bidimensional entropy, each including a number of keyword arguments that allows the user to specify multiple parameters in the entropy calculation. Instructions for installation, descriptions of function syntax, and examples of use are fully detailed in the supporting documentation , available on the EntropyHub website, www.EntropyHub.xyz. Compatible with Windows, Mac and Linux operating systems, EntropyHub is hosted on GitHub, as well as the native package repository for MATLAB, Python and Julia, respectively. The goal of EntropyHub is to integrate the many established entropy methods into one complete resource, providing tools that make advanced entropic time series analysis straightforward and reproducible.
... In recent years, the characteristics of nonlinear dynamics have attracted many scholars' attention because of their advantages in representing nonlinear signals. e commonly used nonlinear dynamics features include Lyapunov index [13], fractal dimension [14], Lempel-Ziv complexity [15,16] and entropy algorithm [17], among which entropy algorithm can be used to represent the amount of information in a period of time, and has been widely used in many fields due to its simplicity of calculation [18,19]. In 2002, Bandt and Pompe first proposed permutation entropy (PE) and applied it to the detection of biomedical signals [20]. ...
Article
Full-text available
The classification and recognition of ship-radiated noise (SRN) is of great significance to the processing of underwater acoustic signals. In order to improve the stability of recognition and more accurately identify SRN, single feature extraction and dual feature extraction based on hierarchical dispersion entropy (HDE) are proposed. For single feature extraction, HDE of the best node among the eight nodes of the third layer decomposition is extracted. For dual feature extraction, HDE of the best two nodes among the 14 nodes of the first-, second-, and third-layer decompositions are required. The results show that the recognition rate of single and dual feature extraction originated from the method based on HDE reaches 85% and 100%, respectively, better than the method of hierarchical reverse dispersion entropy (HRDE) and hierarchical permutation entropy (HPE).
... In fact, the former group complexity calculated with LZ achieved similar results to random noise (meaning high complexity), while in the latter group, its complexity was lower, showing sinusoidal patterns. The applications of compression in health research range from event detection [such as epileptic seizure (86), the onset of ventricular tachycardia or fibrillation (87) and changes from sleep to waking state in-depth anesthesia (88)], characterizing neural spike trains (89), fHR biometric identification (90) or in DNA sequences studies (91). A distinct approach to applying compression on a time series uses the normalized compression distance (NCD) measure, a dissimilarity learning approach first used in fHR by Santos et al. (37). ...
Article
Full-text available
The analysis of fetal heart rate variability has served as a scientific and diagnostic tool to quantify cardiac activity fluctuations, being good indicators of fetal well-being. Many mathematical analyses were proposed to evaluate fetal heart rate variability. We focused on non-linear analysis based on concepts of chaos, fractality, and complexity: entropies, compression, fractal analysis, and wavelets. These methods have been successfully applied in the signal processing phase and increase knowledge about cardiovascular dynamics in healthy and pathological fetuses. This review summarizes those methods and investigates how non-linear measures are related to each paper's research objectives. Of the 388 articles obtained in the PubMed/Medline database and of the 421 articles in the Web of Science database, 270 articles were included in the review after all exclusion criteria were applied. While approximate entropy is the most used method in classification papers, in signal processing, the most used non-linear method was Daubechies wavelets. The top five primary research objectives covered by the selected papers were detection of signal processing, hypoxia, maturation or gestational age, intrauterine growth restriction, and fetal distress. This review shows that non-linear indices can be used to assess numerous prenatal conditions. However, they are not yet applied in clinical practice due to some critical concerns. Some studies show that the combination of several linear and non-linear indices would be ideal for improving the analysis of the fetus's well-being. Future studies should narrow the research question so a meta-analysis could be performed, probing the indices' performance.
... Measurements through the observation of heart rate, breathing pattern, blood pressure and other factors, have been used to measure DoA [6]; however, these physiological signals are the secondary measurements of DoA, showing large crosssubject variability that requires significant experience from anesthesiologists to decode the DoA. In recent years, EEG as an invaluable modality of recording brain activity has attracted more and more attention among anesthesiologists [7]. Compared to the secondary measurements, EEG signal is a direct measurement for brain states as the brain cognitive process relies on communications between neuronal populations through electrical signal [8]. ...
Preprint
Full-text available
Monitoring the depth of unconsciousness during anesthesia is useful in both clinical settings and neuroscience investigations to understand brain mechanisms. Electroencephalogram (EEG) has been used as an objective means of characterizing brain altered arousal and/or cognition states induced by anesthetics in real-time. Different general anesthetics affect cerebral electrical activities in different ways. However, the performance of conventional machine learning models on EEG data is unsatisfactory due to the low Signal to Noise Ratio (SNR) in the EEG signals, especially in the office-based anesthesia EEG setting. Deep learning models have been used widely in the field of Brain Computer Interface (BCI) to perform classification and pattern recognition tasks due to their capability of good generalization and handling noises. Compared to other BCI applications, where deep learning has demonstrated encouraging results, the deep learning approach for classifying different brain consciousness states under anesthesia has been much less investigated. In this paper, we propose a new framework based on meta-learning using deep neural networks, named Anes-MetaNet, to classify brain states under anesthetics. The Anes-MetaNet is composed of Convolutional Neural Networks (CNN) to extract power spectrum features, and a time consequence model based on Long Short-Term Memory (LSTM) Networks to capture the temporal dependencies, and a meta-learning framework to handle large cross-subject variability. We used a multi-stage training paradigm to improve the performance, which is justified by visualizing the high-level feature mapping. Experiments on the office-based anesthesia EEG dataset demonstrate the effectiveness of our proposed Anes-MetaNet by comparison of existing methods.
... It is a simple and powerful method which has been used in several biomedical applications [68]. LZC depends on a coarse-grain processing of the measurements [69] and can be applied directly on physiologic signal without preprocessing [70]. LZC has been applied extensively in analysing biomedical signals (e.g., EEG) to measure the complexity of discrete-time physiologic signals [67]. ...
Preprint
Full-text available
Alzheimers disease (AD) is a progressive disorder that affects cognitive brain functions and starts many years before its clinical manifestations. A biomarker that provides a quantitative measure of changes in the brain due to AD in the early stages would be useful for early diagnosis of AD, but this would involve dealing with large numbers of people because up to 50% of dementia sufferers do not receive a formal diagnosis. Thus, there is a need for accurate, low-cost, and easy-to-use biomarkers that could be used to detect AD in its early stages. Potentially, electroencephalogram (EEG) based biomarkers can play a vital role in early diagnosis of AD as they can fulfill these needs. This is a cross-sectional study that aims to demonstrate the usefulness of EEG complexity measures in early AD diagnosis. We have focused on the three complexity methods which have shown the greatest promise in the detection of AD, Tsallis entropy (TsEn), Higuchi Fractal Dimension (HFD), and Lempel-Ziv complexity (LZC) methods. Unlike previous approaches, in this study, the complexity measures are derived from EEG frequency bands (instead of the entire EEG) as EEG activities have significant association with AD and this has led to enhanced performance. The results show that AD patients have significantly lower TsEn, HFD, and LZC values for specific EEG frequency bands and for specific EEG channels and that this information can be used to detect AD with a sensitivity and specificity of more than 90%.
... In this equation, x t represents the threshold which is often selected as the mean value of the time-series (X.-S. Zhang, Roy, & Jensen, 2001). • Calculate the complexity counter c(N ) representing the total number of distinct patterns/characters contained in the encoded binary string. ...
Article
Full-text available
This article presents an approach to the assessment of operational manufacturing systems complexity based on the irregularities hidden in manufacturing key performance indicator time-series by employing three complementary algorithmic complexity measures: Kolmogorov complexity, Kolmogorov complexity spectrum's highest value and overall Kolmogorov complexity. A series of computer simulations derived from discrete manufacturing systems are used to investigate the measures' potentiality. The results showed that the presented measures can be used in quantitatively identifying operational system complexity, thereby supporting operational shop-floor decision-making activities.
... Anaesthesia is another state associated with low levels of neural complexity (Ferenets et al. 2006(Ferenets et al. , 2007Li et al. 2010;Liang et al. 2015;Schartner et al. 2015;Varley et al. 2020c;Zhang et al. 2001). Like deep NREM sleep, general anaesthesia is often used as a proxy for unconsciousness in studies attempting to validate neural measures of consciousness. ...
Article
Full-text available
Disorders of consciousness (DoCs) pose a significant clinical and ethical challenge because they allow for complex forms of conscious experience in patients where intentional behaviour and communication are highly limited or non-existent. There is a pressing need for brain-based assessments that can precisely and accurately characterize the conscious state of individual DoC patients. There has been an ongoing research effort to develop neural measures of consciousness. However, these measures are challenging to validate not only due to our lack of ground truth about consciousness in many DoC patients but also because there is an open ontological question about consciousness. There is a growing, well-supported view that consciousness is a multidimensional phenomenon that cannot be fully described in terms of the theoretical construct of hierarchical, easily ordered conscious levels. The multidimensional view of consciousness challenges the utility of levels-based neural measures in the context of DoC assessment. To examine how these measures may map onto consciousness as a multidimensional phenomenon, this article will investigate a range of studies where they have been applied in states other than DoC and where more is known about conscious experience. This comparative evidence suggests that measures of conscious level are more sensitive to some dimensions of consciousness than others and cannot be assumed to provide a straightforward hierarchical characterization of conscious states. Elevated levels of brain complexity, for example, are associated with conscious states characterized by a high degree of sensory richness and minimal attentional constraints, but are suboptimal for goal-directed behaviour and external responsiveness. Overall, this comparative analysis indicates that there are currently limitations to the use of these measures as tools to evaluate consciousness as a multidimensional phenomenon and that the relationship between these neural signatures and phenomenology requires closer scrutiny.
... Values nearer to 0 are indicative of a more predictive signal and values approaching 2 are indicative of a more chaotic signal [18]. LZEn involves counting the number of distinct and repeating patterns within a time sequence [19]. Each distinct pattern, or subsequence, is assigned a symbol and then converted to 0 s and 1 s to form a binary substring. ...
Supramaximal interval exercise alters measures of autonomic modulation, while a cool-down may speed the recovery of vagal modulation. The purpose of this study was to compare the effects of a cool-down (pedaling a cycle ergometer at 50 rpm against a resistance of 45 W) versus passive recovery (no pedaling) after supramaximal interval exercise on autonomic modulation. Sixteen moderately active individuals (Mean ± SD: 23 ± 3 years (men: n = 10; women: n = 6) were assessed for autonomic modulation at Rest, and 15 (R15), 30 (R30), 45 (R45) and 60 (R60) min following supramaximal interval exercise. Linear measures of autonomic modulation included natural log (ln) total power (lnTP), high-frequency power (lnHF), the ratio of low frequency (LF) to HF ln(LF/HF) ratio, root mean square of successive differences between normal heartbeats (lnRMSSD), while non-linear measures included sample entropy (SampEn) and Lempel-Ziv entropy (LZEn). Two-way repeated ANOVAs were used to evaluate the main effects of condition (cool-down, passive recovery) across time (Rest, and R15, R30, R45 and R60). There were significant (p ≤ 0.05) condition by time interactions for SampEn and LZEn, such that they decreased at 15, 30, 45 and 60 min during passive recovery compared to Rest, with the recovery of SampEn and LZEn by 60 and 45 min, respectively, during cool-down. There were significant (p ≤ 0.05) main effects of time for lnTP, lnHF and lnRMSSD, such that lnTP, lnHF and lnRMSSD were attenuated, and lnLF/HF ratio was augmented, at all recovery times compared to Rest. These data demonstrate that a cool-down increases the recovery of nonlinear measures of vagal modulation within 45-60 min after supramaximal interval exercise, compared to passive recovery in moderately active individuals.
... In this regard, Gonzalez et al. (2019) have shown that complexity, assessed by permutation entropy, is higher during W than both sleep states . Furthermore, EEG complexity decreases under general anaesthesia (Hudetz et al., 2016;Zhang et al., 2001). Notably, in comparison with W, we also found a decrease in complexity during NREMure and REMure assessed by permutation entropy (data not shown). ...
Article
Urethane is a general anesthetic widely used in animal research. The state of urethane anesthesia is unique because it alternates between macroscopically distinct electrographic states: a slow‐wave state that resembles NREM sleep, and an activated state with features of both REM sleep and wakefulness. Although it is assumed that urethane produces unconsciousness, this has been questioned because of states of cortical activation during drug exposure. Furthermore, the similarities and differences between urethane anesthesia and physiological sleep are still unclear. In this study, we recorded the electroencephalogram (EEG) and electromyogram in chronically prepared rats during natural sleep‐wake states and during urethane anesthesia. We subsequently analyzed the power, coherence, directed connectivity and complexity of brain oscillations, and found that EEG under urethane anesthesia has clear signatures of unconsciousness, with similarities to other general anesthetics. In addition, the EEG profile under urethane is different in comparison to natural sleep states. These results suggest that consciousness is disrupted during urethane. Furthermore, despite similarities that have led others to conclude that urethane is a model of sleep, the electrocortical traits of depressed and activated states during urethane anesthesia differ from physiological sleep states.
... Lempel-Ziv complexity has also been computed from spontaneous neural data, a paradigm that is more readily applicable in clinical and basic research studies as it does not require the use of concurrent TMS or the perturbation of endogenous neural activity. These studies have found constrained complexity in spontaneous neural activity during states marked by low levels of arousal and reductions in conscious contents, such as anesthesia, 70,71,75,78,79,91,92 coma, 82 ...
Thesis
Contemporary theories have argued that the level of consciousness can be approximated by cortical complexity or the strength of frontoparietal connectivity. In support of this, studies have demonstrated suppressed brain state repertoire and frontoparietal connectivity during loss of consciousness. Comparatively little focus has been placed on understanding the neurobiological mechanisms relating to these computational measures of consciousness, leaving their relationship to underlying neurochemical processes unknown. Levels of acetylcholine in the cortex have been shown to relate to the capacity for consciousness, but a relationship between cortical acetylcholine and cortical dynamics such as neurophysiologic complexity has not been investigated. We therefore tested the hypothesis that cortical cholinergic tone would correlate with neurophysiologic complexity. A prior study from our laboratory assessed the effects of cholinergic or noradrenergic stimulation of the prefrontal or parietal cortices in anesthetized rats, finding that only cholinergic stimulation of the prefrontal cortex restored wakefulness during anesthesia. While only prefrontal cholinergic neurotransmission was implicated in regulating the level of consciousness, all stimulation cohorts displayed an activated electroencephalogram (EEG) and elevations in cortical acetylcholine relative to the pre-stimulation anesthetized state. Therefore, in our first data chapter we used EEG data from Pal et al. 2018 to test if prefrontal cholinergic neurotransmission regulates the level of consciousness through changes in neurophysiologic complexity and frontoparietal connectivity. As expected, sevoflurane anesthesia suppressed neurophysiologic complexity and corticocortical connectivity relative to wakefulness. Unexpectedly, however, the strength of frontoparietal connectivity remained suppressed in all cohorts following stimulation, notwithstanding the presence or absence of wakefulness. In contrast, complexity was elevated in all cohorts, correlating instead with spectral features such as EEG activation and periods of elevated cortical acetylcholine. We conclude that prefrontal cholinergic neurotransmission does not regulate the level of consciousness through frontoparietal connectivity, and that neurophysiologic complexity may instead index EEG activation or cortical acetylcholine. In our next data chapter, we explored the relationship between EEG complexity, spectral contents of the signal, and the level of consciousness by contrasting the effects of ketamine and propofol anesthesia over three bandwidths (0.5-175 Hz, 65-175 Hz, and 0.5-55 Hz). We demonstrate bandwidth-dependent properties of both ketamine and propofol on complexity, demonstrating that ketamine anesthesia suppresses complexity in the 65-175 Hz bandwidth, while propofol does not. Using a normalization method to average out spectral influence on the signal, we demonstrate comparable effects of ketamine and propofol anesthesia on 0.5-175 Hz and 0.5-55 Hz complexity that are dissociable from spectral EEG properties. In our final data chapter, we leveraged the dose-dependent properties of the dissociative anesthetics ketamine and nitrous oxide to characterize the relationship between cortical acetylcholine levels, neurophysiologic complexity, and frontoparietal connectivity (25-55 Hz, 85-125 Hz, 125-175 Hz). During subanesthetic ketamine and nitrous oxide induction, we report periods of elevated prefrontal and parietal cortical cholinergic tone, neurophysiologic complexity, and frontoparietal connectivity in the 125-175 Hz bandwidth. During nitrous oxide sedation, cortical acetylcholine levels and neurophysiologic complexity were concomitantly suppressed with the level of arousal. Our findings demonstrate a correlation between cortical acetylcholine levels, neurophysiologic complexity, and frontoparietal connectivity in the high gamma bandwidth. In sum, our findings establish acetylcholine as a neurochemical correlate of cortical dynamics purported to relate to consciousness. While future causal studies are necessary, we offer preliminary evidence suggesting a role for cortical cholinergic neurotransmission in supporting neurophysiologic complexity and the capacity for consciousness.
... Lempel-Ziv complexity (LZ) is a measure of the diver- sequence. When applied to neuroimaging data, lower LZ (with respect to wakeful rest) has been associated with unconscious states such as sleep [35] or anaesthesia [36], and higher LZ with states of richer phenomenal content under psychedelics, ketamine [11,12] and states of flow during musical improvisation [37]. ...
Preprint
Full-text available
Schizophrenia and states induced by certain psychotomimetic drugs may share some physiological and phenomenological properties, but they differ in fundamental ways: one is a crippling chronic mental disease, while the others are temporary, pharmacologically-induced states presently being explored as treatments for mental illnesses. Building towards a deeper understanding of these different alterations of normal consciousness, here we compare the changes in neural dynamics induced by LSD and ketamine (in healthy volunteers) against those associated with schizophrenia, as observed in resting-state M/EEG recordings. While both conditions exhibit increased neural signal diversity, our findings reveal that this is accompanied by an increased transfer entropy from the front to the back of the brain in schizophrenia, versus an overall reduction under the two drugs. Furthermore, we show that these effects can be reproduced via different alterations of standard Bayesian inference applied on a computational model based on the predictive processing framework. In particular, the effects observed under the drugs are modelled as a reduction of the precision of the priors, while the effects of schizophrenia correspond to an increased precision of sensory information. These findings shed new light on the similarities and differences between schizophrenia and two psychotomimetic drug states, and have potential implications for the study of consciousness and future mental health treatments.
Preprint
Full-text available
There is increasing evidence that level of consciousness can be captured by neural informational complexity: for instance, complexity, as measured by the Lempel Ziv (LZ) compression algorithm, decreases during anesthesia and non-rapid eye movement (NREM) sleep in humans and rats, when compared to LZ in awake and REM sleep. In contrast, LZ is higher in humans under the effect of psychedelics, including subanesthetic doses of ketamine. However, it is both unclear how this result would be modulated by varying ketamine doses, and whether it would extend to other species. Here we studied LZ with and without auditory stimulation during wakefulness and different sleep stages in 5 cats implanted with intracranial electrodes, as well as under subanesthetic doses of ketamine (5, 10, and 15 mg/kg i.m.). In line with previous results, LZ was lowest in NREM sleep, but similar in REM and wakefulness. Furthermore, we found an inverted U-shaped curve following different levels of ketamine doses in a subset of electrodes, primarily in prefrontal cortex. However, it is worth noting that the variability in the ketamine dose-response curve across cats and cortices was larger than that in the sleep-stage data, highlighting the differential local dynamics created by two different ways of modulating conscious state. These results replicate previous findings, both in humans and other species, demonstrating that neural complexity is highly sensitive to capture state changes between wake and sleep stages while adding a local cortical description. Finally, this study describes the differential effects of ketamine doses, replicating a rise in complexity for low doses, and further fall as doses approach anesthetic levels in a differential manner depending on the cortex.
Book
This book, based on a selection of invited presentations from a topical workshop, focusses on time-variable oscillations and their interactions. The problem is challenging, because the origin of the time variability is usually unknown. In mathematical terms, the oscillations are nonautonomous, reflecting the physics of open systems where the function of each oscillator is affected by its environment. Time-frequency analysis being essential, recent advances in this area, including wavelet phase coherence analysis and nonlinear mode decomposition, are discussed. Some applications to biology and physiology are described. Although the most important manifestation of time-variable oscillations is arguably in biology, they also crop up in, e.g. astrophysics, or for electrons on superfluid helium. The book brings together the research of the best international experts in seemingly very different disciplinary areas.
Article
Objective We demonstrate that multifrequency entropy gives insight into the relationship between epileptogenicity and sleep, and forms the basis for an improved measure of medical assessment of sleep impairment in epilepsy patients. Methods Multifrequency entropy was computed from electroencephalography measurements taken from 31 children with Benign Epilepsy with Centrotemporal Spikes and 31 non-epileptic controls while awake and during sleep. Values were compared in the epileptic zone and away from the epileptic zone in various sleep stages. Results We find that I) in lower frequencies, multifrequency entropy decreases during non-rapid eye movement sleep stages when compared with wakefulness in a general population of pediatric patients, II) patients with Benign Epilepsy with Centrotemporal Spikes had lower multifrequency entropy across stages of sleep and wakefulness, and III) the epileptic regions of the brain exhibit lower multifrequency entropy patterns than the rest of the brain in epilepsy patients. Conclusions Our results show that multifrequency entropy decreases during sleep, particularly sleep stage 2, confirming, in a pediatric population, an association between sleep, lower multifrequency entropy, and increased likelihood of seizure. Significance We observed a correlation between lowered multifrequency entropy and increased epileptogenicity that lays preliminary groundwork for the detection of a digital biomarker for epileptogenicity.
Article
The condition monitoring(CM) of rotating machinery(RM) is an essential operation for improving the reliability of mechanical systems. For this purpose, an efficient CM method that possesses simple and intuitive attributes is required for industrial applications. For condition monitoring that connects fault detection, degradation assessment, and prognosis applications, health indicators(HIs) have been developed in the past few decades. The construction of a HI is the decisive procedure for extracting informative fault information from the monitoring signal. From the initial statistical parameter-based construction methods to the introduction of data-oriented intelligent methods such as deep learning in recent years, HIs construction methods have ranged from fault mechanism-based approaches to a data-based approach, which involve two different technologies regardless of superiority or inferiority. This paper provides a systematic review of the HIs construction methods for rotating machinery proposed in the literature. It emphasizes the classical technical approaches and recent interesting research trends and analyzes the benefits and potential of efficient HIs for condition monitoring. The current challenges and future research opportunities are also presented in this paper. The Engineers and researchers interested in this research can be informed of current research ideas and directions in the field by reading this paper, as well as inspiring potentially excellent research work in the future.
Article
This work proposed a novel method for automatic sleep stage classification based on the time, frequency, and fractional Fourier transform (FRFT) domain features extracted from a single-channel electroencephalogram (EEG). Bidirectional long short-term memory was applied to the proposed model to train it to learn the sleep stage transition rules according to the American Academy of Sleep Medicine's manual for automatic sleep stage classification. Results indicated that the features extracted from the fractional Fourier-transformed single-channel EEG may improve the performance of sleep stage classification. For the Fpz-Cz EEG of Sleep-EDF with 30 s epochs, the overall accuracy of the model increased by circa 1% with the help of the FRFT domain features and even reached 81.6%. This work thus made the application of FRFT to automatic sleep stage classification possible. The parameters of the proposed model measured 0.31 MB, which are 5% of those of DeepSleepNet, but its performance is similar to that of DeepSleepNet. Hence, the proposed model is a light and efficient model based on deep neural networks, which also has a prospect for on-device machine learning.
Article
Full-text available
The brain is universally regarded as a system for processing information. If so, any behavioral or cognitive dysfunction should lend itself to depiction in terms of information processing deficiencies. Information is characterized by recursive, hierarchical complexity. The brain accommodates this complexity by a hierarchy of large/slow and small/fast spatiotemporal loops of activity. Thus, successful information processing hinges upon tightly regulating the spatiotemporal makeup of activity, to optimally match the underlying multiscale delay structure of such hierarchical networks. Reduced capacity for information processing will then be expressed as deviance from this requisite multiscale character of spatiotemporal activity. This deviance is captured by a general family of multiscale criticality measures (MsCr). MsCr measures reflect the behavior of conventional criticality measures (such as the branching parameter) across temporal scale. We applied MsCr to MEG and EEG data in several telling degraded information processing scenarios. Consistently with our previous modeling work, MsCr measures systematically varied with information processing capacity: MsCr fingerprints showed deviance in the four states of compromised information processing examined in this study, disorders of consciousness, mild cognitive impairment, schizophrenia and even during pre-ictal activity. MsCr measures might thus be able to serve as general gauges of information processing capacity and, therefore, as normative measures of brain health.
Article
Evidence suggests that the stream of consciousness is parsed into transient brain states manifesting themselves as discrete spatiotemporal patterns of global neuronal activity. Electroencephalographical (EEG) microstates are proposed as the neurophysiological correlates of these transiently stable brain states that last for fractions of seconds. To further understand the link between EEG microstate dynamics and consciousness, we continuously recorded high-density EEG in 23 surgical patients from their awake state to unconsciousness, induced by step-wise increasing concentrations of the intravenous anesthetic propofol. Besides the conventional parameters of microstate dynamics, we introduce a new implementation of a method to estimate the complexity of microstate sequences. The brain activity under the surgical anesthesia showed a decreased sequence complexity of the stereotypical microstates, which became sparser and longer-lasting. However, we observed an initial increase in microstates’ temporal dynamics and complexity with increasing depth of sedation leading to a distinctive "U-shape" that may be linked to the paradoxical excitation induced by moderate levels of propofol. Our results support the idea that the brain is in a metastable state under normal conditions, balancing between order and chaos in order to flexibly switch from one state to another. The temporal dynamics of EEG microstates indicate changes of this critical balance between stability and transition that lead to altered states of consciousness.
Preprint
Full-text available
Background: Electroencephalography (EEG) is the most common method to access brain information. Techniques to monitor and extract brain signal characteristics in farm animals are not as developed as in humans and laboratory animals. New method: The method comprised two steps. In the first step, the signals were acquired after the telemetric equipment was developed, the electrodes were positioned and fixed, the sampling frequency was defined, the equipment was positioned, and artifacts and other acquisition problems were dealt with. Brain signals from six Holstein heifers that could move freely in free stalls were acquired. The control group consisted of the same number of bovines, contained in a climatic chamber (restrained group). In the second step, the signals were characterized by Power Spectral Density, Short-Time Fourier Transform, and Lempel-Ziv complexity. Results: The results indicated that there was an ideal position to attach the electrodes to the front of the bovine's head so that longer artifact-free signal sections were acquired. The signals showed typical EEG frequency bands, like the bands found in humans. The Lempel-Ziv complexity values indicated that the bovine brain signals contained random and chaotic components. As expected, the signals acquired from the retained bovine group displayed sections with a larger number of artifacts. Comparison with existing methods: We present the first method that helps to monitor and to extract brain signal features in unrestrained bovines. Conclusions: The method could be applied to investigate changes in brain electrical activity during animal farming, to monitor brain activity related to animal behavior.
Article
Real-time emotion recognition with electroencephalograph (EEG) has been an active field of research in recent years. In particular, deep learning has been shown to be effective in emotion classification tasks. However, the monitoring of EEG signals is a continuous process, there is a need for energy-efficient emotion classification methods. Compared with artificial neural networks (ANNs), spiking neural networks (SNNs), in which weight multiplications are replaced by additions, are more energy efficient. In this paper, we propose a near-lossless transfer learning method for SNNs, specially designed for EEG signals. Data is preprocessed, and its power spectral density (PSD) is extracted to represent the frequency domain of the raw EEG signal. Using a 3-layer pretrained SNN, running on the DEAP dataset, we achieved an accuracy of 78.87% and 76.5% for valence and arousal dimensions, respectively. By training a model based on one dimension and fine-tuning on another, we even achieve higher accuracy, 82.75% for the valence and 84.22% for the arousal. As far as we know, our results yield the smallest SNN with the highest accuracy for this task to date. The energy power of our SNNs for valence and arousal dimensions is 13.8% that of our CNN-based solutions. The framework was developed by PyTorch and is available under an open-source license.
Article
Full-text available
Over the last years, a surge of empirical studies converged on complexity-related measures as reliable markers of consciousness across many different conditions, such as sleep, anesthesia, hallucinatory states, coma, and related disorders. Most of these measures were independently proposed by researchers endorsing disparate frameworks and employing different methods and techniques. Since this body of evidence has not been systematically reviewed and coherently organized so far, this positive trend has remained somewhat below the radar. The aim of this paper is to make this consilience of evidence in the science of consciousness explicit. We start with a systematic assessment of the growing literature on complexity-related measures and identify their common denominator, tracing it back to core theoretical principles and predictions put forward more than 20 years ago. In doing this, we highlight a consistent trajectory spanning two decades of consciousness research and provide a provisional taxonomy of the present literature. Finally, we consider all of the above as a positive ground to approach new questions and devise future experiments that may help consolidate and further develop a promising field where empirical research on consciousness appears to have, so far, naturally converged.
Conference Paper
Polysomnography (PSG), the gold standard for sleep stage classification, requires a sleep expert for scoring and is both resource-intensive and expensive. Many researchers currently focus on the real-time classification of the sleep stages based on biomedical signals, such as Electroencephalograph (EEG) and electrooculography (EOG). However, most of the research work is based on machine learning models with multiple signal inputs or hand-engineered features requiring prior knowledge of the sleep domain. We propose a novel encoded TimeDistributed Convolutional Neural Network (TDConvNet) to automatically classify sleep stages based on a single raw PSG signal. The TDConvNet can infer sleep stages in just 30-second epochs for the 5-stage sleep classification using a single EEG or EOG signal from the open Sleep-EDF dataset. We evaluated the TDConvNet performance on EEGs and EOG signals. The evaluation results show that the proposed method achieved the best performance with the EEG Fpz-Cz signal (0.85) compared to current literature, followed by EEG Pz-Oz (0.84) and EOG horizontal (0.82). The source code is available at https://github.com/michealdutt/TDConvNet.
Article
Measuring complexity from noisy time series provides a crucial insight into the understanding and monitoring of pattern dynamics of complex dynamical systems, including physiological systems such as complex electroencephalogram (EEG) times series observed during anesthesia. We introduce a simple, yet noble complexity measure, called the lumped permutation entropy (LPE) based on the permutation entropy (PE), which allows a tie rank on the pattern formation and shows the robustness under the influence of strong noise, overcoming some limitations of PE and its variants in noisy signals. The robustness of LPE is demonstrated for complex time series from a typical chaotic dynamical system and is applied to empirical electroencephalographic (EEG) data obtained from subjects anesthetized by propofol. In particular, we found that the entropic complexity of EEG time series based on LPE is inversely correlated with plasma concentration of propofol and shows better performance and more robustness than PE and other types of entropic algorithms in indicating the anesthetic depth during the progress of general anesthesia. LPE can be used as complexity measure for real-time monitoring of anesthetic depth during anesthesia.
Article
Related experiments have shown that transcranial direct current stimulation (tDCS) anodal stimulation of the brain’s primary motor cortex (M1) and supplementary motor area (SMA) can improve the motor control and clinical manifestations of stroke patients with aphasia and dyskinesia. In this study, to explore the different effects of tDCS on the M1 and SMA in motor imagery, 35 healthy volunteers participated in a double-blind randomized controlled experiment. Five subjects underwent sham stimulation (control), 15 subjects underwent tDCS anode stimulation of the M1, and the remaining 15 subjects underwent tDCS anode stimulation of the SMA. The electroencephalogram data of the subjects’ left- and right-hand motor imagery under different stimulation paradigms were recorded. We used a functional brain network and sample entropy to examine the different complexities and functional connectivities in subjects undergoing sham-tDCS and the two stimulation paradigms. The results show that tDCS anodal stimulation of the SMA produces less obvious differences in the motor preparation phase, while tDCS anodal stimulation of the M1 produces significant differences during the motor imaging task execution phase. The effect of tDCS on the motor area of the brain is significant, especially in the M1.
Article
Full-text available
We review several aspects of the analysis of time sequences, and concentrate on recent methods using concepts from the theory of nonlinear dynamical systems. In particular, we discuss problems in estimating attractor dimensions, entropies, and Lyapunov exponents, in reducing noise and in forecasting. For completeness and since we want to stress connections to more traditional (mostly spectrum-based) methods, we also give a short review of spectral methods.
Article
Full-text available
We have previously derived a theoretical measure of neural complexity (CN) in an attempt to characterize functional connectivity in the brain. CN measures the amount and heterogeneity of statistical correlations within a neural system in terms of the mutual information between subsets of its units. CN was initially used to characterize the functional connectivity of a neural system isolated from the environment. In the present paper, we introduce a related statistical measure, matching complexity (CM), which reflects the change in CN that occurs after a neural system receives signals from the environment. CM measures how well the ensemble of intrinsic correlations within a neural system fits the statistical structure of the sensory input. We show that CM is low when the intrinsic connectivity of a simulated cortical area is randomly organized. Conversely, CM is high when the intrinsic connectivity is modified so as to differentially amplify those intrinsic correlations that happen to be enhanced by sensory input. When the input is represented by an individual stimulus, a positive value of CM indicates that the limited mutual information between sensory sheets sampling the stimulus and the rest of the brain triggers a large increase in the mutual information between many functionally specialized subsets within the brain. In this way, a complex brain can deal with context and go "beyond the information given."
Article
Full-text available
Traditional feature extraction methods describe signals in terms of amplitude and frequency. This paper takes a paradigm shift and investigates four stochastic-complexity features. Their advantages are demonstrated on synthetic and physiological signals; the latter recorded during periods of Cheyne-Stokes respiration, anesthesia, sleep, and motor-cortex investigation.
Article
Full-text available
Conventional approaches to understanding consciousness are generally concerned with the contribution of specific brain areas or groups of neurons. By contrast, it is considered here what kinds of neural processes can account for key properties of conscious experience. Applying measures of neural integration and complexity, together with an analysis of extensive neurological data, leads to a testable proposal-the dynamic core hypothesis-about the properties of the neural substrate of consciousness.
Article
Full-text available
We investigate several quantifiers of the electroencephalogram (EEG) signal with respect to their ability to indicate depth of anesthesia. For 17 patients anesthetized with sevoflurane, three established measures (two spectral and one based on the bispectrum), as well as a phase space based nonlinear correlation index were computed from consecutive EEG epochs. In the absence of an independent way to determine anesthesia depth, the standard was derived from measured blood plasma concentrations of the anesthetic via a pharmacokinetic/pharmacodynamic model for the estimated effective brain concentration of sevoflurane. In most patients, the highest correlation is observed for the nonlinear correlation index D*. In contrast to spectral measures, D* is found to decrease monotonically with increasing (estimated) depth of anesthesia, even when a "burst-suppression" pattern occurs in the EEG. The findings show the potential for applications of concepts derived from the theory of nonlinear dynamics, even if little can be assumed about the process under investigation.
Article
Full-text available
Techniques to determine changing system complexity from data are evaluated. Convergence of a frequently used correlation dimension algorithm to a finite value does not necessarily imply an underlying deterministic model or chaos. Analysis of a recently developed family of formulas and statistics, approximate entropy (ApEn), suggests that ApEn can classify complex systems, given at least 1000 data values in diverse settings that include both deterministic chaotic and stochastic processes. The capability to discern changing complexity from such a relatively small amount of data holds promise for applications of ApEn in a variety of contexts.
Article
Full-text available
Approximate entropy (ApEn) is a recently developed statistic quantifying regularity and complexity, which appears to have potential application to a wide variety of relatively short (greater than 100 points) and noisy time-series data. The development of ApEn was motivated by data length constraints commonly encountered, e.g., in heart rate, EEG, and endocrine hormone secretion data sets. We describe ApEn implementation and interpretation, indicating its utility to distinguish correlated stochastic processes, and composite deterministic/ stochastic models. We discuss the key technical idea that motivates ApEn, that one need not fully reconstruct an attractor to discriminate in a statistically valid manner-marginal probability distributions often suffice for this purpose. Finally, we discuss why algorithms to compute, e.g., correlation dimension and the Kolmogorov-Sinai (KS) entropy, often work well for true dynamical systems, yet sometimes operationally confound for general models, with the aid of visual representations of reconstructed dynamics for two contrasting processes. (c) 1995 American Institute of Physics.
Article
We studied the associated factors and incidence of awareness during general anesthesia and the nature of subsequent psychiatric disorders.Patients older than 12 yr undergoing surgery under general anesthesia in a secondary care hospital during 1 yr were included in the study. The doses of anesthetics were calculated for the patients with and without awareness. There were 4818 operations under general anesthesia; 2612 (54%) patients were interviewed. Ten (0.4% of those interviewed) patients were found to have undisputed awareness, and there were nine (0.3%) patients with possible awareness. The doses of isoflurane (P < 0.01) and propofol (P < 0.05) were smaller in patients with awareness. Five patients with awareness underwent a psychiatric evaluation. One patient experienced sleep disturbances afterward, but the other four patients did not have any after effects. In conclusion, awareness is a rare complication of general anesthesia associated with small doses of anesthetics. Implications: In an interview of 2612 patients after general anesthesia, 10 (0.4%) patients with awareness and 9 (0.3%) patients with possible awareness were found. A predisposing factor was small doses of the principal anesthetic. In a psychiatric interview, a large proportion of the patients with awareness were found to have suffered from depression in the past. (Anesth Analg 1998;86:1084-9)
Article
A new measure of strange attractors is introduced which offers a practical algorithm to determine their character from the time series of a single observable. The relation of this new measure to fractal dimension and information-theoretic entropy is discussed.
Article
The measure of dimensional complexity has the potential for feature extraction, modeling and prediction of EEG signals. However, the nonlinear dynamics of neuronal processes is under criticism that EEG signals may have a simpler stochastic description and chaotic dynamical measures of EEG may be spurious or unnecessary. Surrogate-data testing has been propounded to detect nonlinearity and chaos in experimental time series and to differentiate it from linear stochastic processes or colored noises. The surrogate data tests of brain signals (EEG) have produced equivocal results. Therefore, we examine the surrogate testing procedure using numerical data of classical chaotic systems, mixed sine waves, white Gaussian and colored Gaussian noises and typical EEGs. White Gaussian noise and classical chaotic time series are easily discerned by the surrogate-data test. However, a colored Gaussian noise data of low correlation dimensions (D2) or mixed sine waves containing less number of sinusoids show behaviors similar to the low dimensional deterministic chaotic systems. There are significant differences in D2 values between the original and surrogate data sets. The colored Gaussian noise appears linear and stochastic only when there is an increased randomness in its pattern and the signal is high dimensional. Our results clearly indicate that the "surrogate testing" alone may not be a sufficient test for distinguishing colored noises from low dimensional chaos. The EEG time series produce finite correlation dimensions. The surrogate testing of 8 independent realizations of different forms of EEG activities produce significantly different D2 values than the original data sets. Apparently many natural phenomena follow deterministic chaos and as the dimensional complexity of the system increases (D2 > 5) it may approximate a stochastic process. Thus EEG appears unlikely to have originated from a linear system driven by white noise.
Article
A popular procedure for testing a pattern recognition machine is to present the machine with a set of patterns taken from the real world. The proportion of these patterns which are misrecognized or rejected is taken as the estimate of the error probability or rejection probability for the machine. In Part I, this testing procedure is discussed for the cases of unknown and known a priori probabilities of occurrence of the pattern classes. The differences between the tests that should be made in the two cases are noted, and confidence intervals for the test results are indicated. These concepts are applied to various published pattern recognition results by determining the appropriate confidence interval for each result. In Part II, the problem of the optimum partitioning of a sample of fixed size between the design and test phases of a pattern recognition machine is discussed. One important nonparametric result is that the proportion of the total sample used for testing the machine should never be less than that proportion used for designing the machine, and in some cases should be a good deal more.
Article
The Observer's Assessment of Alertness/Sedation (OAA/S) Scale was developed to measure the level of alertness in subjects who are sedated. This scale was tested in 18 subjects in a three-period crossover study to assess its reliability and its criterion, behavioral, and construct validity. After receiving either placebo or a titrated dose of midazolam to produce light or heavy sedation, each subject was administered two sedation scales (OAA/S Scale and a Visual Analogue Scale) and two performances tests (Digit Symbol Substitution Test and Serial Sevens Subtraction). Two raters individually evaluated the subject's level of alertness on each of the two sedation scales. The results obtained on the OAA/S Scale were reliable and valid as measured by high correlations between the two raters and high correlations between the OAA/S Scale and two of the three standard tests used in this study. The OAA/S Scale was sensitive to the level of midazolam administered; all pairwise comparisons were significant (p < 0.05) for all three treatment levels at both test periods. (C) Williams & Wilkins 1990. All Rights Reserved.
Article
The electroencephalogram (EEG) was used because of its dimensional complexity to establish a differentiation of divergent versus convergent thought, considered fundamental modes of cortical processing. In 28 men, the EEG was recorded while solving tasks of divergent and convergent thinking and during mental relaxation. The EEG during divergent thought was compared between subjects achieving high versus low performance scores on this type of task. The dimensional complexity of the EEG was greater during divergent thinking than during convergent thinking. While solving tasks of divergent thinking, subjects with high performance scores had a lower EEG dimension than did subjects with low scores, in particular over frontal cortical areas. The changes were not reflected in single frequency bands of conventional EEG analysis. Based on Hebb's view of neuron assemblies as functional processing units, the higher EEG complexity during divergent than convergent thinking could be the result of the concurrent activation of a greater number of independently oscillating processing units.
Article
This paper gives the main results of our studies of information transmission in the human cerebral cortex. The analysis is based on the mutual information theory of Vastano and Swinney (1988). We recorded eight-channel electroencephalographs (EEGs). The information transmission between these eight leads was computed. The transmission intensities were characterized by complexity measures. The complexity measures are: the algorithm complexity of Kolmogorov (1965), C1 complexity and C2 complexity. The last two were defined by us in a previous paper (Xu et al., 1994).
Article
Dimensional complexity (DCx) is an EEG measure derived from nonlinear systems theory that can be indicative of the global dynamical complexity of electrocortical activity. This study examined developmental changes in DCx, as well as the effects of cognitive tasks, gender, and brain topography, and compared DCx with traditional spectral power measures. EEG was recorded in three groups of children at mean age of 7.5 (n = 37), 13.8 (n = 42), and 16.4 (n = 56) years at rest and during the performance of verbal and spatial cognitive tasks. DCx measured both at rest and during tasks increased with age. Specific effects of brain topography, condition, and gender became stronger with age, suggesting an increase in structural and functional differentiation of the cortex. Hemispheric asymmetry of DCx recorded during tasks also increased with age, with the task-induced DCx reduction being stronger in the left hemisphere. Gender differences in DCx suggested faster cerebral maturation in girls over late adolescence. Relationships between DCx and spectral power varied as a function of tasks and scalp locations, suggesting that these EEG measures can reflect different aspects of cortical functioning.
Article
The present study examined EEG dimensional complexity (estimated correlation dimension) in 76 healthy volunteers in response to emotionally valenced (i.e. neutral, positive and negative) film clip stimulation. EEG was recorded from 18 sites (10-20 system). We estimated the dimensional complexity by the Grassberger and Procaccia and Skinner's point-wise dimension (PD2i) methods. The results were compared to spectral measures of the EEG. Only the PD2i algorithm (i.e. the one that did not require data stationarity) discriminated among all the three film categories. The main results showed that both negative and positive emotions occurred with higher values (at some posterior locations) of EEG DCx estimates compared to the neutral viewing condition. The topographical differences (frontal vs. posterior temporal) between positive and negative evoked emotions were obtained. There were also some significant direct relationships between dynamic complexity estimates and intensities of subjective emotional feelings. It is concluded that dimensional complexity estimates turned out to be sensitive to subtle aspects of emotional processing not accessible by linear EEG analyses.
Article
Five numerical descriptors were derived from the electroencephalogram (EEG), recorded, and processed (Tracor Nomad) during emergence from isoflurane-nitrous oxide anesthesia. The five descriptors (median frequency, spectral edge frequency-90%, total power, a frequency band power ratio, and the ratio of frontal to occipital power) were compared for their ability to predict imminent arousal. Arousal was defined as spontaneous movement, coughing or eye opening. All of the descriptors except the frontal-occipital power ratio underwent significant (P less than 0.05) changes between the initial recordings made intraoperatively during surgical stimulus under anesthesia and later recordings in the 40 s preceding arousal. A post hoc analysis was performed to identify the threshold value for each parameter that best served to predict imminent arousal. For median frequency, spectral edge frequency-90%, total power, and the frequency band power ratio, thresholds that predicted imminent arousal with sensitivities of 90% and specificities of 82-90% could be identified. The data indicate that, even in the favorable circumstances of the present study (uniform anesthetic technique, post hoC identification of thresholds), none of several previously popularized EEG descriptors (median frequency, spectral edge frequency-90%, total power, a frequency band power ratio) can serve as a completely reliable sole predictor of imminent arousal. As presently derived, these EEG descriptors at best provide trend information to be used in concert with other clinical signs of depth of anesthesia.
Article
Changes in brain activity were studied at different depths of isoflurane anaesthesia. Ten healthy women (ASA group l) were investigated during non-critical surgery. Two channels of the EEG were stored on tape simultaneously with alveolar concentration of carbon dioxide, inspired oxygen concentration, mean arterial pressure, ECG and temperature. Signal processing was made off-line. Spectral information from 2-s EEG segments was extracted using autoregressive modelling. Repetitive hierarchical clustering was used to define a common learning set of basic patterns. With this learning set, the EEG was classified, and the results presented in a class probability histogram. The basic patterns were related to the clinical depth of anaesthesia in all patients and assigned specific colours. Using this colour code, the class probability histogram showed a high degree of simplicity. Decreasing or increasing the isoflurane concentration caused the same trend in the class profile in all patients. This indicates that the EEG pattern might be a sensitive tool for decision making during administration of general anaesthetics.
Intraoperative monitoring of electroencephalography (EEG) data can help assess brain integrity and/or depth of anesthesia. We demonstrate a computer generated technique which provides a visually robust display of EEG data plotted as ‘phase space trajectories’ and a mathematically derived parameter (‘dimensionality’) which may correlate with depth of anesthesia. Application of nonlinear mathematical analysis, used to describe complex dynamical systems, can characterize ‘phase space’ EEG patterns by identifying attractors (geometrical patterns in phase space corresponding to specific ordered EEG data subjects) and by quantifying the degree of order and chaos (calculation of dimensionality). Dimensionality calculations describe the degree of complexity in a signal and may generate a clinically useful univariate EEG descriptor of anesthetic depth. In this paper we describe and demonstrate phase space trajectories generated for sine waves, mixtures of sine waves, and white noise (random chaotic events). We also present EEG phase space trajectories and dimensionality calculations from a patient undergoing surgery and general anesthesia in 3 recognizable states: awake, anesthetized, and burst suppression. Phase space trajectories of the three states are visually distinguishable, and dimensionality calculations indicate that EEG progresses from ‘chaos’ (awake) to progressively more ‘ordered’ attractors (anesthetized and burst suppression).
Article
The degree of interaction of component waves making up a single electroencephalogram trace was strongly correlated with alpha activity, lead placement, and state of consciousness. Significant quadratic coupling of the waves was found only for awake subjects with high alpha activity. For these subjects about 50 percent of beta activity can be attributed to harmonic coupling with the alpha peak. During sleep, the degree of interaction was of borderline significance and did not follow a consistent pattern with respect to subject, frequency, state, or lead.
Article
We recorded the electroencephalogram (EEG) in 16 patients during propofol/sufentanil total intravenous anesthesia to determine whether EEG changes might predict imminent awakening during emergence. Changes in absolute and relative power in four frequency bands, median frequency (MF), 95th percentile frequency (F95), and two frequency band power ratios (beta/alpha and (alpha+beta)/delta) were quantified. One minute before eye opening, absolute power in the delta and alpha bands had decreased to 49% (25%-73%) and 42% (25%-58%) of the value during the infusion (P > 0.005). MF, F95, and the two frequency band power ratios increased during emergence (P > 0.05). Of the individual spectral variables, only a 50% decrease in absolute alpha power was more than 90% sensitive and specific in predicting eye opening. We conclude that, although pronounced EEG changes occur during emergence from propofol/sufentanil anesthesia, the EEG does not reliably predict eye opening.
Article
The objective of our study was to test the efficacy of the bispectral index (BIS) compared with spectral edge frequency (SEF), relative delta power, median frequency, and a combined univariate power spectral derivative in predicting movement to incision during isoflurane/oxygen anesthesia. A total of 42 consenting patients were assigned to 3 groups, isoflurane 0.75, 1.0, and 1.25 minimal alveolar concentration (MAC). Anesthesia was induced with thiopental and maintained with the appropriate end-tidal concentration of isoflurane. The electroencephalogram (EEG) was recorded using a microcomputer system, and data were analyzed off-line. The EEG during the 2 min before incision was analyzed. Following skin incision, each patient was carefully observed for 60 sec to detect occurrence of purposeful movement. For all groups combined, there was a statistically significant difference for BIS (p < 0.0001) and also for relative delta power (p < 0.016) between movers and nonmovers. There was a statistically significant difference between movers and nonmovers at 1.25 MAC isoflurane for BIS (p < 0.01). There were no other significant differences for any other EEG variable at any concentration of isoflurane. No EEG variable showed a relationship to isoflurane concentration. When bispectral analysis of the EEG was used to develop a retrospectively determined index, there was an association of the index with movement. Thus, it may be a useful predictor of whether patients will move in response to skin incision during anesthesia with isoflurane/oxygen.
Article
A time-frequency spectral representation (TFSR) has been used to study the nonstationary information in the EEG as an aid in determining the anesthetic depth. This paper uses a TFSR with an exponential weighting function for the purpose. Raw EEG data were collected form 10 mongrel dogs at various levels of halothane anesthesia. Depth of anesthesia was tested by observing the response to tail clamping, which is considered a supramaximal stimulus in dogs. A positive response was graded as awake (depth 0), and a negative response was graded as asleep (depth 1). The EEG obtained during a period of 30 sec tail clamp was processed into TFSRs. It was observed that at depth 0, the spectrum becomes localized in time and frequency. The percentage of energy in the delta (1-3.5 Hz) and theta (3.5-7.5 Hz) frequency bands increased. At depth 1, the spectrum remained unchanged throughout the period of tail clamp. The performance of the TFSR in detecting the patient's awareness was also compared with the power spectrum. It was concluded that under certain anesthetic conditions, the TFSR is superior to the power spectrum.
Analysis of the EEG as a signal from a deterministic non-linear system should, in principle, allow insights into the complexity of underlying brain activity. We examined the capability of this method to analyse the marked changes in brain activity during normal brain development. Resting EEGs of 54 healthy children (newborns to 14 years old) and of 12 normal adults were recorded digitally. The following parameters were calculated: correlation dimension, a measure of the complexity of the underlying system, and the first Lyapunov coefficient, indicating the system's 'unpredictability'. Analysis of variance (ANOVA) was performed with probands grouped by age. The subgroups of children older than 1 year was further examined by regression analysis. In all analysed epochs, Lyapunov coefficients were significantly positive (P < 0.0001. t-test). The presence of non-linear dynamics was asserted statistically in 64-76% of examined epochs. A highly significant increase in correlation dimension with age was found in all examined leads (P < 0.0001, ANOVA). In all age groups, marked differences in correlation dimension in different brain regions became evident (P < 0.01-0.0001, ANOVA). Evidence for the presence of non-linearity can be found even in newborns. Brain maturation was reflected in a marked and highly significant increase in correlation dimension (complexity). Our work indicates that non-linear dynamics analysis is suitable for measuring complexity of brain activity during maturation and provides age-dependent normal values as a basis for further study.
Article
Fractal dimensions has been proposed as a useful measure for the characterisation of electrophysiological time series. But one of the problems of this approach, is the difficulty to record time series long enough of determine the 'real' fractal dimension. Nevertheless it is possible to calculate fractal dimensions for very short data-segments. Using time series of different length it is possible to show, that there is a monotoneous relation between fractal dimension and the number of data-points. This relation could be further interpreted with the help of an extrapolation scheme. In addition this effect is also seen with surrogate data, generated from that signal. We conclude that it is feasible to use fractal dimension as a tool to characterise the complexity for short electroencephalographic (EEG) time series, but it is not possible to decide whether the brain is a chaotic system or not.
Article
The need for a reliable method of predicting movement during anesthesia has existed since the introduction of anesthesia. This paper proposes a recognition system, based on the autoregressive (AR) modeling and neural network analysis of the electroencephalograph (EEG) signals, to predict movement following surgical stimulation. The input to the neural network will be the AR parameters, the hemodynamic parameters blood pressure (BP) and heart rate (HR), and the anesthetic concentration in terms of the minimum alveolar concentration (MAC). The output will be the prediction of movement. Design of the system and results from the preliminary tests on dogs are presented in this paper. The experiments were carried out on 13 dogs at different levels of halothane. Movement prediction was tested by monitoring the response to tail clamping, which is considered to be a supramaximal stimulus in dogs. The EEG data obtained prior to tail clamping was processed using a tenth-order AR model and the parameters obtained were used as input to a three-layer perceptron feedforward neural network. Using only AR parameters the network was able to correctly classify subsequent movement in 85% of the cases as compared to 65% when only hemodynamic parameters were used as the input to the network. When both the measures were combined, the recognition rate rose to greater than 92%. When the anesthetic concentration was added as an input the network could be considerably simplified without sacrificing classification accuracy. This recognition system shows the feasibility of using the EEG signals for movement during anesthesia.