Unsupervised Multitaper Spectral Method for Identifying REM Sleep in Intracranial EEG Recordings Lacking EOG/EMG Data.

A large number of human intracranial EEG (iEEG) recordings have been collected for clinical purposes, in institutions all over the world, but the vast majority of these are unaccompanied by EOG and EMG recordings which are required to separate Wake episodes from REM sleep using accepted methods. In order to make full use of this extremely valuable data, an accurate method of classifying sleep from iEEG recordings alone is required. Existing methods of sleep scoring using only iEEG recordings accurately classify all stages of sleep, with the exception that wake (W) and rapid-eye movement (REM) sleep are not well distinguished. A novel multitaper (Wake vs. REM) alpha-rhythm classifier is developed by generalizing K-means clustering for use with multitaper spectral eigencoefficients. The performance of this unsupervised method is assessed on eight subjects exhibiting normal sleep architecture in a hold-out analysis and is compared against a classical power detector. The proposed multitaper classifier correctly identifies 36±6 min of REM in one night of recorded sleep, while incorrectly labeling less than 10% of all labeled 30 s epochs for all but one subject (human rater reliability is estimated to be near 80%), and outperforms the equivalent statistical-power classical test. Hold-out analysis indicates that when using one night's worth of data, an accurate generalization of the method on new data is likely. For the purpose of studying sleep, the introduced multitaper alpha-rhythm classifier further paves the way to making available a large quantity of otherwise unusable IEEG data.

Emotional Memory Processing during REM Sleep with Implications for Post-Traumatic Stress Disorder.

REM sleep is important for the processing of emotional memories, including fear memories. Rhythmic interactions, especially in the theta band, between the medial prefrontal cortex (mPFC) and limbic structures are thought to play an important role, but the ways in which memory processing occurs at a mechanistic and circuits level are largely unknown. To investigate how rhythmic interactions lead to fear extinction during REM sleep, we used a biophysically based model that included the infralimbic cortex (IL), a part of the mPFC with a critical role in suppressing fear memories. Theta frequency (4-12 Hz) inputs to a given cell assembly in IL, representing an emotional memory, resulted in the strengthening of connections from the IL to the amygdala and the weakening of connections from the amygdala to the IL, resulting in the suppression of the activity of fear expression cells for the associated memory. Lower frequency (4 Hz) theta inputs effected these changes over a wider range of input strengths. In contrast, inputs at other frequencies were ineffective at causing these synaptic changes and did not suppress fear memories. Under post-traumatic stress disorder (PTSD) REM sleep conditions, rhythmic activity dissipated, and 4 Hz theta inputs to IL were ineffective, but higher-frequency (10 Hz) theta inputs to IL induced changes similar to those seen with 4 Hz inputs under normal REM sleep conditions, resulting in the suppression of fear expression cells. These results suggest why PTSD patients may repeatedly experience the same emotionally charged dreams and suggest potential neuromodulatory therapies for the amelioration of PTSD symptoms.SIGNIFICANCE STATEMENT Rhythmic interactions in the theta band between the mPFC and limbic structures are thought to play an important role in processing emotional memories, including fear memories, during REM sleep. The infralimbic cortex (IL) in the mPFC is thought to play a critical role in suppressing fear memories. We show that theta inputs to the IL, unlike other frequency inputs, are effective in producing synaptic changes that suppress the activity of fear expression cells associated with a given memory. Under PTSD REM sleep conditions, lower-frequency (4 Hz) theta inputs to the IL do not suppress the activity of fear expression cells associated with the given memory but, surprisingly, 10 Hz inputs do. These results suggest potential neuromodulatory therapies for PTSD.

Cortical circuit-based lossless neural integrator for perceptual decision-making: A computational modeling study.

The intrinsic uncertainty of sensory information (i.e., evidence) does not necessarily deter an observer from making a reliable decision. Indeed, uncertainty can be reduced by integrating (accumulating) incoming sensory evidence. It is widely thought that this accumulation is instantiated via recurrent rate-code neural networks. Yet, these networks do not fully explain important aspects of perceptual decision-making, such as a subject's ability to retain accumulated evidence during temporal gaps in the sensory evidence. Here, we utilized computational models to show that cortical circuits can switch flexibly between "retention" and "integration" modes during perceptual decision-making. Further, we found that, depending on how the sensory evidence was readout, we could simulate "stepping" and "ramping" activity patterns, which may be analogous to those seen in different studies of decision-making in the primate parietal cortex. This finding may reconcile these previous empirical studies because it suggests these two activity patterns emerge from the same mechanism.

Multitaper estimates of phase-amplitude coupling.

Phase-amplitude coupling (PAC) is the association of the amplitude of a high-frequency oscillation with the phase of a low-frequency oscillation. In neuroscience, this relationship provides a mechanism by which neural activity might be coordinated between distant regions. The dangers and pitfalls of assessing PAC with commonly used statistical measures have been well-documented. The limitations of these measures include: (1) response to non-oscillatory, high-frequency, broad-band activity, (2) response to high-frequency components of the low-frequency oscillation, (3) adhoc selection of analysis frequency-intervals, and (4) reliance upon data shuffling to assess statistical significance.Objective.To address issues (1)-(4) by introducing a nonparametric multitaper estimator of PAC.Approach.In this work, a multitaper PAC estimator is proposed that addresses these issues. Specifically, issue (1) is addressed by replacing the analytic signal envelope estimator computed using the Hilbert transform with a multitaper estimator that down-weights non-sinusoidal activity using a classical, multitaper super-resolution technique. Issue (2) is addressed by replacing coherence between the low-frequency and high-frequency components in a standard PAC estimator with multitaper partial coherence, while issue (3) is addressed with a physical argument regarding meaningful neural oscillation. Finally, asymptotic statistical assessment of the multitaper estimator is introduced to address issue (4).Main results.Multitaper estimates of PAC are introduced. Their efficacy is demonstrated in simulation and on human intracranial recordings obtained from epileptic patients.Significance.This work facilitates a more informative statistical assessment of PAC, a phenomena exhibited by many neural systems, and provides a basis upon which further nonparametric multitaper-related methods can be developed.

Spatially expandable fiber-based probes as a multifunctional deep brain interface.

Understanding the cytoarchitecture and wiring of the brain requires improved methods to record and stimulate large groups of neurons with cellular specificity. This requires miniaturized neural interfaces that integrate into brain tissue without altering its properties. Existing neural interface technologies have been shown to provide high-resolution electrophysiological recording with high signal-to-noise ratio. However, with single implantation, the physical properties of these devices limit their access to one, small brain region. To overcome this limitation, we developed a platform that provides three-dimensional coverage of brain tissue through multisite multifunctional fiber-based neural probes guided in a helical scaffold. Chronic recordings from the spatially expandable fiber probes demonstrate the ability of these fiber probes capturing brain activities with a single-unit resolution for long observation times. Furthermore, using Thy1-ChR2-YFP mice we demonstrate the application of our probes in simultaneous recording and optical/chemical modulation of brain activities across distant regions. Similarly, varying electrographic brain activities from different brain regions were detected by our customizable probes in a mouse model of epilepsy, suggesting the potential of using these probes for the investigation of brain disorders such as epilepsy. Ultimately, this technique enables three-dimensional manipulation and mapping of brain activities across distant regions in the deep brain with minimal tissue damage, which can bring new insights for deciphering complex brain functions and dynamics in the near future.

Frontal beta-theta network during REM sleep.

We lack detailed knowledge about the spatio-temporal physiological signatures of REM sleep, especially in humans. By analyzing intracranial electrode data from humans, we demonstrate for the first time that there are prominent beta (15-35 Hz) and theta (4-8 Hz) oscillations in both the anterior cingulate cortex (ACC) and the DLPFC during REM sleep. We further show that these theta and beta activities in the ACC and the DLPFC, two relatively distant but reciprocally connected regions, are coherent. These findings suggest that, counter to current prevailing thought, the DLPFC is active during REM sleep and likely interacting with other areas. Since the DLPFC and the ACC are implicated in memory and emotional regulation, and the ACC has motor areas and is thought to be important for error detection, the dialogue between these two areas could play a role in the regulation of emotions and in procedural motor and emotional memory consolidation.

Thalamic mechanisms underlying alpha-delta sleep with implications for fibromyalgia.

Alpha-delta sleep is the abnormal intrusion of alpha activity (8- to 13-Hz oscillations) into the delta activity (1- to 4-Hz oscillations) that defines slow-wave sleep. Alpha-delta sleep is especially prevalent in fibromyalgia patients, and there is evidence suggesting that the irregularities in the sleep of these patients may cause the muscle and tissue pain that characterizes the disorder. We constructed a biophysically realistic mathematical model of alpha-delta sleep. Imaging studies in fibromyalgia patients suggesting altered levels of activity in the thalamus motivated a thalamic model as the source of alpha activity. Since sodium oxybate helps to alleviate the symptoms of fibromyalgia and reduces the amount of alpha-delta sleep in fibromyalgia patients, we examined how changes in the molecular targets of sodium oxybate affected alpha-delta activity in our circuit. Our model shows how alterations in GABAB currents and two thalamic currents, Ih (a hyperpolarization-activated current) and a potassium leak current, transform a circuit that normally produces delta oscillations into one that produces alpha-delta activity. Our findings suggest that drugs that reduce Ih conductances and/or increase potassium conductances, without necessarily increasing GABAB conductances, might be sufficient to restore delta sleep. Furthermore, they suggest that delta sleep might be restored by drugs that preferentially target these currents in the thalamus; such drugs might have fewer side effects than drugs that act systemically.

A time-series model of phase amplitude cross frequency coupling and comparison of spectral characteristics with neural data.

Stochastic processes that exhibit cross-frequency coupling (CFC) are introduced. The ability of these processes to model observed CFC in neural recordings is investigated by comparison with published spectra. One of the proposed models, based on multiplying a pulsatile function of a low-frequency oscillation (θ) with an unobserved and high-frequency component, yields a process with a spectrum that is consistent with observation. Other models, such as those employing a biphasic pulsatile function of a low-frequency oscillation, are demonstrated to be less suitable. We introduce the full stochastic process time series model as a summation of three component weak-sense stationary (WSS) processes, namely, θ, γ, and η, with η a 1/f (α) noise process. The γ process is constructed as a product of a latent and unobserved high-frequency process x with a function of the lagged, low-frequency oscillatory component (θ). After demonstrating that the model process is WSS, an appropriate method of simulation is introduced based upon the WSS property. This work may be of interest to researchers seeking to connect inhibitory and excitatory dynamics directly to observation in a model that accounts for known temporal dependence or to researchers seeking to examine what can occur in a multiplicative time-domain CFC mechanism.

Brain rhythms connect impaired inhibition to altered cognition in schizophrenia.

In recent years, schizophrenia research has focused on inhibitory interneuron dysfunction at the level of neurobiology and on cognitive impairments at the psychological level. Reviewing both experimental and computational findings, we show how the temporal structure of the activity of neuronal populations, exemplified by brain rhythms, can begin to bridge these levels of complexity. Oscillations in neuronal activity tie the pathophysiology of schizophrenia to alterations in local processing and large-scale coordination, and these alterations in turn can lead to the cognitive and perceptual disturbances observed in schizophrenia.

Thalamocortical mechanisms for the anteriorization of α rhythms during propofol-induced unconsciousness.

As humans are induced into a state of general anesthesia via propofol, the normal alpha rhythm (8-13 Hz) in the occipital cortex disappears and a frontal alpha rhythm emerges. This spatial shift in alpha activity is called anteriorization. We present a thalamocortical model that suggests mechanisms underlying anteriorization. Our model captures the neural dynamics of anteriorization when we adjust it to reflect two key actions of propofol: its potentiation of GABA and its reduction of the hyperpolarization-activated current Ih. The reduction in Ih abolishes the occipital alpha by silencing a specialized subset of thalamocortical cells, thought to generate occipital alpha at depolarized membrane potentials (>-60 mV). The increase in GABA inhibition imposes an alpha timescale on both the cortical and thalamic portions of the frontal component that are reinforced by reciprocal thalamocortical feedback. Anteriorization can thus be understood as a differential effect of anesthetic drugs on thalamic nuclei with disparate spatial projections, i.e.: (1) they disrupt the normal, depolarized alpha in posterior-projecting thalamic nuclei while (2) they engage a new, hyperpolarized alpha in frontothalamic nuclei. Our model generalizes to other anesthetics that include GABA as a target, since the molecular targets of many such anesthetics alter the model dynamics in a manner similar to that of propofol.

Biophysical Modeling of Alpha Rhythms During Halothane-Induced Unconsciousness.

During the induction of general anesthesia there is a shift in power from the posterior regions of the brain to the frontal cortices; this shift in power is called anteriorization. For many anesthetics, a prominent feature of anteriorization is a shift specifically in the alpha band (8-13 Hz) from posterior to frontal cortices. Here we present a biophysical computational model that describes thalamocortical circuit-level dynamics underlying anteriorization of the alpha rhythm in the case of halothane. Halothane potentiates GABAA and increases potassium leak conductances. According to our model, an increase in potassium leak conductances hyperpolarizes and silences the high-threshold thalamocortical (HTC) cells, a specialized subset of thalamocortical cells that fire at the alpha frequency at relatively depolarized membrane potentials (>-60 mV) and are thought to be the generators of quiet awake occipital alpha. At the same time the potentiation of GABAA imposes an alpha time scale on both the cortical and the thalamic component of the frontal portion of our model. The alpha activity in the frontal component is further strengthened by reciprocal thalamocortical feedback. Thus, we argue that the dual molecular targets of halothane induce the anteriorization of the alpha rhythm by increasing potassium leak conductances, which abolishes occipital alpha, and by potentiating GABAA, which induces frontal alpha. These results provide a computational modeling formulation for studying highly detailed biophysical mechanisms of anesthetic action in silico.

Thalamic model of awake alpha oscillations and implications for stimulus processing.

We describe a unique conductance-based model of awake thalamic alpha and some of its implications for function. The full model includes a model for a specialized class of high-threshold thalamocortical cells (HTC cells), which burst at the alpha frequency at depolarized membrane potentials (~-56 mV). Our model generates alpha activity when the actions of either muscarinic acetylcholine receptor (mAChR) or metabotropic glutamate receptor 1 (mGluR1) agonists on thalamic reticular (RE), thalamocortical (TC), and HTC cells are mimicked. In our model of mGluR1-induced alpha, TC cells are equally likely to fire during any phase of alpha, consistent with in vitro experiments. By contrast, in our model of mAChR-induced alpha, TC cells tend to fire either at the peak or the trough of alpha, depending on conditions. Our modeling suggests that low levels of mGluR1 activation on a background of mAChR agonists may be able to initiate alpha activity that biases TC cells to fire at certain phases of alpha, offering a pathway for cortical control. If we introduce a strong stimulus by increasing the frequency of excitatory postsynaptic potentials (EPSPs) to TC cells, an increase in alpha power is needed to mimic the level of phasing of TC cells observed in vivo. This increased alpha power reduces the probability that TC cells spike near the trough of alpha. We suggest that mAChR-induced alpha may contribute to grouping TC activity into discrete perceptual units for processing, whereas mGluR1-induced alpha may serve the purpose of blocking unwanted stimuli from reaching the cortex.

A neurophysiological-metabolic model for burst suppression.

Burst suppression is an electroencepholagram (EEG) pattern in which high-voltage activity alternates with isoelectric quiescence. It is characteristic of an inactivated brain and is commonly observed at deep levels of general anesthesia, hypothermia, and in pathological conditions such as coma and early infantile encephalopathy. We propose a unifying mechanism for burst suppression that accounts for all of these conditions. By constructing a biophysical computational model, we show how the prevailing features of burst suppression may arise through the interaction between neuronal dynamics and brain metabolism. In each condition, the model suggests that a decrease in cerebral metabolic rate, coupled with the stabilizing properties of ATP-gated potassium channels, leads to the characteristic epochs of suppression. Consequently, the model makes a number of specific predictions of experimental and clinical relevance.

Assessing neuronal interactions of cell assemblies during general anesthesia.

Understanding the way in which groups of cortical neurons change their individual and mutual firing activity during the induction of general anesthesia may improve the safe usage of many anesthetic agents. Assessing neuronal interactions within cell assemblies during anesthesia may be useful for understanding the neural mechanisms of general anesthesia. Here, a point process generalized linear model (PPGLM) was applied to infer the functional connectivity of neuronal ensembles during both baseline and anesthesia, in which neuronal firing rates and network connectivity might change dramatically. A hierarchical Bayesian modeling approach combined with a variational Bayes (VB) algorithm is used for statistical inference. The effectiveness of our approach is evaluated with synthetic spike train data drawn from small and medium-size networks (consisting of up to 200 neurons), which are simulated using biophysical voltage-gated conductance models. We further apply the analysis to experimental spike train data recorded from rats' barrel cortex during both active behavior and isoflurane anesthesia conditions. Our results suggest that that neuronal interactions of both putative excitatory and inhibitory connections are reduced after the induction of isoflurane anesthesia.

Activity in the barrel cortex during active behavior and sleep.

The rate at which neurons fire has wide-reaching implications for the coding schemes used by neural systems. Despite the extensive use of the barrel cortex as a model system, relatively few studies have examined the rate of sensory activity in single neurons in freely moving animals. We examined the activity of barrel cortex neurons in behaving animals during sensory cue interaction, during non-stimulus-related activity, during various states of sleep, and during the administration of isoflurane. The activity of regular-spiking units (RSUs: predominantly excitatory neurons) and fast spiking units (FSUs: a subtype of inhibitory interneurons) was examined separately. We characterized activity by calculating neural firing rates, because several reports have emphasized the low firing rates in this system, reporting that both baseline activity and stimulus evoked activity is <1 Hz. We report that, during sensory cue interaction or non-stimulus-related activity, the majority of RSUs in rat barrel cortex fired at rates significantly >1 Hz, with 27.4% showing rates above 10 Hz during cue interaction. Even during slow wave sleep, which had the lowest mean and median firing rates of any nonanesthetized state observed, 80.0% of RSUs fired above 1 Hz. During all of the nonanesthetized states observed 100% of the FSUs fired well above 1 Hz. When rats were administered isoflurane and at a depth of anesthesia used in standard in vivo electrophysiological preparations, all of the RSUs fired below 1 Hz. We also found that >80% of RSUs either upmodulated or downmodulated their firing during cue interaction. These data suggest that low firing rates do not typify the output of the barrel cortex during awake activity and during sleep and indicate that sensory coding at both the individual and population levels may be nonsparse.

Discrete- and continuous-time probabilistic models and algorithms for inferring neuronal UP and DOWN states.

UP and DOWN states, the periodic fluctuations between increased and decreased spiking activity of a neuronal population, are a fundamental feature of cortical circuits. Understanding UP-DOWN state dynamics is important for understanding how these circuits represent and transmit information in the brain. To date, limited work has been done on characterizing the stochastic properties of UP-DOWN state dynamics. We present a set of Markov and semi-Markov discrete- and continuous-time probability models for estimating UP and DOWN states from multiunit neural spiking activity. We model multiunit neural spiking activity as a stochastic point process, modulated by the hidden (UP and DOWN) states and the ensemble spiking history. We estimate jointly the hidden states and the model parameters by maximum likelihood using an expectation-maximization (EM) algorithm and a Monte Carlo EM algorithm that uses reversible-jump Markov chain Monte Carlo sampling in the E-step. We apply our models and algorithms in the analysis of both simulated multiunit spiking activity and actual multi- unit spiking activity recorded from primary somatosensory cortex in a behaving rat during slow-wave sleep. Our approach provides a statistical characterization of UP-DOWN state dynamics that can serve as a basis for verifying and refining mechanistic descriptions of this process.

Vascular responses to syntactic processing: event-related fMRI study of relative clauses.

Event-related functional magnetic resonance imaging was used to investigate the localization of syntactic processing in sentence comprehension. Matched pairs of sentences containing identical lexical items were compared. One member of the pair consisted of a syntactically simpler sentence, containing a subject relativized clause. The second member of the pair consisted of a syntactically more complex sentence, containing an object relativized clause. Ten subjects made plausibility judgments about the sentences, which were presented one word at a time on a computer screen. There was an increase in BOLD hemodynamic signal in response to the presentation of all sentences compared to fixation in both right and left occipital cortex, the left perisylvian cortex, and the left premotor and motor areas. BOLD signal increased in the left angular gyrus when subjects processed the complex portion of syntactically more complex sentences. This study shows that a hemodynamic response associated with processing the syntactically complex portions of a sentence can be localized to one part of the dominant perisylvian association cortex.