EEG Sleep Spindles Research Articles

Procedural replay: The anatomy and physics of the sleep spindle

This commentary implicates the neostriatum in the production of the EEG sleep spindle and in the processing of motor procedural learning in sleep. Whether the sleep spindle may implement not only the consolidation-based enhancement of procedural learning, but also its initial consolidation, is considered; as is the fit between (1) corticostriatal anatomy and physiology, and (2) the physical properties of the sleep spindle.

Automatic recognition of sleep spindles in EEG via radial basis support vector machine based on a modified feature selection algorithm

This paper presents an application of a radial basis support vector machine (RB-SVM) to the recognition of the sleep spindles (SSs) in electroencephalographic (EEG) signal. The proposed system comprises of two stages. In the first stage, for feature extraction, a set of raw amplitude values, a set of discrete cosine transform (DCT) coefficients, a set of discrete wavelet transform (DWT) approximation coefficients and a set of adaptive autoregressive (AAR) parameters are calculated and extracted from signals separately as four different sets of feature vectors. Thus, four different feature vectors for the same data are comparatively examined. In the second stage, these features are then selected by a modified adaptive feature selection method based on sensitivity analysis, which mainly supports input dimension reduction via selecting the most significant feature elements. Then, the feature vectors are classified by a support vector machine (SVM) classifier, which is relatively new and powerful technique for solving supervised binary classification problems due to it’s generalization ability. Visual evaluation, by two electroencephalographers (EEGers), of 19 channel EEG records of six subjects showed that the best performance is obtained with an RB-SVM providing an average sensitivity of 97.7%, an average specificity of 97.4% and an average accuracy of 97.5%.

Automatic recognition of sleep spindles in EEG by using artificial neural networks

In this paper, we introduce a two-stage procedure based on artificial neural networks for the automatic recognition of sleep spindles (SSs) in a multi-channel electroencephalographic signal. In the first stage, a discrete perceptron is used to eliminate definite non-SSs. The pre-classification done in the first stage not only reduces the computation time but also increases the overall detection performance of the procedure. In the second stage, the remaining SS candidates after pre-classification procedure are aimed to be separated from each other by an artificial neural network that would function as a post-classifier. Two different networks, i.e. a backpropagation multilayer perceptron and radial basis support vector machine (SVM), are proposed as the post-classifier and compared in terms of their classification performances. Visual evaluation, by two electroencephalographers (EEGers), of 19 channel EEG records of 6 subjects showed that the best performance is obtained with a radial basis SVM providing an average sensitivity of 94.6% and an average false detection rate of 4.0%.

Characteristics of human EEG sleep spindles assessed by Gabor transform

The aim of this study is to show an application of the Gabor transform on the detection and characterization of human sleep EEG spindles in a sample of 10 healthy young adults, trying to identify the most useful parameters that can be used for the automatic detection and characterization of such events.

Sleep spindles in human prefrontal cortex: an electrocorticographic study

To investigate the sleep spindle activity of the human prefrontal cortex (PFC), we simultaneously recorded whole nights of polysomnographic and electrocorticographic (ECoG) activities during the natural sleep of epileptic patients. Subjects were nine patients with intractable epilepsy who had subdural electrodes surgically attached to the orbital (seven cases), medial (three cases), or dorsolateral (two cases) PFC, and in one case to the frontal pole. To examine spindle frequencies, fast Fourier transformation (FFT) and auto-correlation analyses were performed on the PFC ECoG and Cz EEG data, primarily on epochs of stage 2 sleep. Lower sigma band ECoG oscillations of about 12 Hz were widely distributed across all prefrontal cortical areas including the frontal limbic regions, but none of the PFC sigma frequency peaks coincided with the faster (about 14 Hz) Cz EEG sleep spindles. Combining our results with anatomical and electrophysiological facts, it is suggested that the thalamofrontal circuit involving the rostral reticular and the mediodorsal nucleus of the thalamus is responsible for the generation of 12 Hz frontal spindles in humans.

Walking through a maze alters the architecture of sleep.

Human and animal studies imply that sleep is a critical period for consolidation of recent memories. Whereas the majority of researchers focussed on the procedural learning, the present human study concerns how storing of spatial information and episodic memory are linked to sleep stages. Two city mazes, a simple and a complex one, were created by means of a computer program. Local aspects of these mazes appeared as street scenes on a TV-screen. Our subjects sat in front of the screen and manoeuvered through the maze by the help of a three-button PC mouse. Thus, each subject took a 'mental walk' through an imaginary city. The task was to find various end-points and to find the way back to the starting point. Subjects of two experimental groups 'walked' through either the simple or complex city maze for eight hours. Afterwards the subjects slept in our laboratory, where their sleep stages could be measured polygraphically. Subjects who had explored the simple maze showed considerable alteration in sleep architecture. They remained significantly longer in sleep Stage 2 than subjects who had explored the complex maze. Moreover, with successful orientation in the simple maze sleep stages occurred aperiodically, whereas walking through the complex maze was associated with sleep stages in accordance with ultradian cycles, as observed in a control group. Compared to subjects of the control group who had experienced neither maze, the subjects of both experimental groups had significantly enhanced EEG sleep spindle activities. Alteration in temporal architecture of sleep and selective prolongation of sleep Stage 2 following spatial orientation point to a functional linkage between cognitive mapping of space and sleep Stage 2 with enhanced EEG spindle activity.

A laboratory study of sleep and dreaming in a case of Asperger's Syndrome.

Asperger's Syndrome (AS) is a pervasive developmental disorder whose continuity with High-Functioning Autism is still a matter of debate. Clinical observations suggest that patients with AS may present the same sleep disorders as autistic patients, including difficulties in initiating and maintaining sleep as well as poor dream recall. We recorded the sleep of a 25-year-old male patient with AS for two nights using a full EEG montage and compared the second night to that of a group of normal participants. We found low levels of slow wave sleep (SWS: stages 3 + 4), high levels of stage 1, and a large number of awakenings. The organization of REM sleep was unremarkable, including normal REM density. Analyses of phasic EEG events revealed a very low incidence of sleep spindles and a normal number of K-complexes over bilateral frontal and central EEG leads. In order to collect dream reports, the patient was awakened three times over two nights following at least 15 minutes of REM sleep in each case. On each occasion the patient was not aware of any mental activity happening just prior to awakening. These observations are discussed with regards to the connections that may exist between EEG sleep spindle activity, selective attention, and the capacity to generate a dream report.

The Mechanism of Sleep Arousal in Infants: Relevance of Sleep Startles• 1834

Defective arousal mechanisms are viewed as contributory to sleep hypopnea disorders and Sudden Infant Death Syndrome. Arousal has traditionally been viewed as an “all or none” event defined behaviorally as“full arousal” (eyes open, cry) or electrophysiologically as >3 sec of altered EEG-activity. In contrast, our own recent studies suggest that arousal is a stepwise process beginning with a sleep startle which then may or may not progress to “full arousal”. However startles have not traditionally been viewed as arousal related phenomena in infants since the associated EEG changes last < 3sec and are mostly due to movement artifacts. We hypothesized that, if startles are the initial event in a sequential arousal process, then they might be associated with less obvious electrophysiological events intermediate between those of sleep and“full arousal”. We monitored EEG sleep spindles since activation of the primary arousal mechanisms in the brainstem reticular activating system eliminates spindle activity generated in the thalamus. We studied spontaneous sleep startles (n=23) as well as those elicited by briefly occluding the infants face mask airway (n=128) in 12 normal infants (age 10-19 weeks) during sleep. We recorded EEG (C3-P3), ECG, O2-Saturation, limb and nuchal EMG and video of the infant. The intensity of startles was scored on a scale of 0 to 3 based on the limb EMG magnitude and video recorded movements. We focused on the β-spindle frequency range (13-18 Hz). Spindle periodicity was analysed by using a threshold over a compressed spectral band array. Spontaneous sleep startles were immediately followed by a prolongation of the inter spindle interval from a mean of 7.9 sec (+/- 0.16 SE) during the control period to 20.4 sec (+/- 0.94 SE) after startle (p<0.005); following occlusion related startles inter spindle period increased to 24.8 sec (+/- 0.92 SE) (p<0.001). Furthermore the intensity of evoked startles was positively correlated with the inter spindle interval period (p<0.01). Conclusion: Spontaneous as well as evoked startles are immediately followed by elimination of sleep spindles. This phenomenon strongly suggests a linkage between sleep startles and increased activity in subcortical arousal mechanisms. These findings support the concept that infant arousal is not an“all or none” event but a process that begins as a series of subcortical events and reflex behaviors that potentially progress to a“full” or unambiguous EEG arousal.HD 10993

Improved procedure of complex demodulation and an application to frequency analysis of sleep spindles in EEG.

Complex demodulation is a local version of harmonic analysis that enables the amplitude and phase of particular frequency components of a time series to be described as functions of time. The paper presents a computational procedure involving complex demodulation with interpolation of data in the frequency domain. A computational procedure comprising repeated use of complex demodulation is also presented. This is used to estimate the optimum choice of the demodulating frequency which considerably influences the measurement of the instantaneous amplitude and phase of the underlying process. The usefulness of this procedure is verified by computer simulation. An example of applying this procedure to the estimation of the centre and the instantaneous frequencies of sleep spindles in the EEG (electroencephalogram) is presented. By using the procedure developed here, several partially overlapping sleep spindles are detected and correctly separated. The paper also presents an approach to separating and analysing transient time series (such as overlapping sleep spindles) by using an accurate frequency processing technique.

Changes in sensorimotor sleep spindle activity and seizure susceptibility after somatosensory deafferentation

This investigation examined the effects of dorsal column transections on susceptibility to monomethylhydrazine (MMH)-induced seizures and sensorimotor EEG sleep spindle activity in the cat. Eleven animals, prepared with indwelling cortical electrodes over hind limb and forelimb projection areas of somatosensory cortex, received bilateral lesions of the dorsal columns at either high (C 1–C 3) or low (C 5-T 1) cervical spinal levels. After a 3-week recovery period, each animal was administered a convulsive dose of MMH. Seizure susceptibility was measured in terms of latency to onset of tonic-clonic convulsions. Sensorimotor 12- to 15-Hz sleep spindle activity was assessed in comparable segments of the pre- and postlesion EEG. Results showed a progressive and significant decrease in seizure susceptibility with successively more encephalad dorsal column transections. The EEG data showed significant postlesion enhancements of sleep spindle activity over cortical sites roughly corresponding to peripheral receptive fields below the level of the lesions. Results suggested an association between the extent of spindle enhancement and the magnitude of MMH seizure response attenuation. This finding was consistent with previous reports indicating a relationship between rhythmic 12- to 15-Hz sensorimotor EEG activity and central inhibitory processes.

Quantitative analysis of infant EEG development during quiet sleep

A quantitative evaluation of central cortical EEG activity during quiet sleep was undertaken in normal infants followed from 1 to 24 weeks of age. Bilateral (C 3-T 3, C 4-T 4) EEG activity was recorded continuously on magnetic tape during 12-h polygraphic monitoring sessions at 1, 4, 8, 12, 16 and 24 weeks of age. Power-spectral densities were calculated from three independent 10 min quiet sleep samples from the first, middle and last epochs of the night. Data were calibrated, log transformed and sorted into five sequential 4 c/sec frequency bands between 0–19 c/sec. Analysis focused upon the distribution of power in these bands as a function of time of night, hemisphere and age. A time of night effect was found only between 8 and 16 weeks of age. During this period the initial sleep epoch showed a significant increase in low frequency power. This was interpreted as an age-specific sleep deprivation effect related to experimental manipulation of the infant. No significant developmental difference in power distribution was found between hemispheres, although specific asymmetries were noted. All frequency bands changed significantly with age. Lower frequencies (0–3 and 4–7 c/sec) increased in power with age, 0–3 c/sec increasing abruptly from 1 to 8 weeks and 4–7 c/sec showing a more delayed increment after 12 weeks of age. Power in the higher frequency bands (8–11, 12–15 and 16–19 c/sec decreased between 1 and 4 weeks and then increased significantly at 12 weeks and older. The decrease at 4 weeks and increase at 12 weeks was most marked for the 12–15 c/sec band, and was interpreted in terms of the development of thalamocortical mechanisms related to the generation of EEG sleep spindles.

Detection of human sleep EEG waveforms

A method for detecting waveforms occurring in the human sleep EEG (alpha, beta, delta, sleep spindles, and muscle activity) is presented. Design criteria are provided for the waveform detectors, which can be readily constructed from linear and digital integrated circuits or programmed on a minicomputer.

The effect of extrauterine experience on the ontogenesis of EEG sleep spindles.

The development of EEG sleep-spindles in prematurely-born and full-term infants was studied to test an hypothesized impact of experience (longer extrauterine existence) upon innate maturational factors (EEG development). The age of onset of sleep spindles and achievement of mature spindle quality oc

SLEEP CYCLE DURING NAPPING.

The cyclic occurrence of REM periods (sleep characterized by a low voltage, desynchronized EEG pattern, and rapid eye movements) 1,3 and the association of these periods with reports of dreaming 1,4,6,7,9,10,14 have been well established. With the exception of narcoleptic sleep, 5,13 and the sleep of amphetamine addicts subjected to drug withdrawal, 12 REM periods do not occur during nocturnal sleep until after the passage of approximately 45 minutes of NREM sleep (periods characterized by EEG sleep spindle and/or δ-activity and the absence of rapid eye movements). Although reports of dreaming are sometimes elicited on awakenings from NREM sleep, the frequency of such reports is less than the frequency for REM period awakenings, and the quality of these reports is less dream-like than the quality of REM period reports. 6,14 Therefore, one would expect, upon hearing vivid, well-recalled, thematically organized dreams, that they were experienced during

INTERRELATEDNESS OF MENTAL ACTIVITY DURING SLEEP.

A paper by Aserinsky and Kleitman 1 in 1953 initiated renewed interest in the scientific study of dreams. They reported that subjects related dreams on 74% of awakenings from periods of sleep characterized by a low-voltage, desynchronized EEG pattern and rapid eye movements (REM periods). The subjects rarely reported dreams when awakened from periods of sleep characterized by EEG delta and sleep spindle activity and no rapid eye movements (NREM periods). A more extensive study by Dement and Kleitman 3 yielded 80% dream recall on REM awakenings and 7% dream recall on NREM awakenings. They considered that the small amount of recall obtained on NREM awakenings probably represented dreams experienced during earlier REM periods and incorrectly assigned by the subject to the NREM period immediately preceding the awakening. Using the REM period as the criterion of dreaming, Dement and Kleitman 2 established that three