[Sleep spindles-Function, detection and use as biomarker for diagnostics in psychiatry]. / Schlafspindeln ­ Funktion, Detektion und Nutzung als Biomarker für die psychiatrische Diagnostik.

BACKGROUND: The sleep spindle is a graphoelement of an electroencephalogram (EEG), which can be observed in light and deep sleep. Alterations in spindle activity have been described for a range of psychiatric disorders. Due to their relatively constant properties, sleep spindles may therefore be potential biomarkers in psychiatric diagnostics. METHOD: This article presents an overview of the state of the science on the characteristics and functions of the sleep spindle as well as known alterations of spindle activity in psychiatric disorders. Various methodological approaches and developments of spindle detection are explained with respect to their potential for application in psychiatric diagnostics. RESULTS AND CONCLUSION: Although alterations in spindle activity in psychiatric disorders are known and have been described in detail, their exact potential for psychiatric diagnostics has yet to be fully determined. In this respect, the acquisition of knowledge in research is currently constrained by manual and automated methods for spindle detection, which require high levels of resources and are error prone. Newer approaches to spindle detection based on deep-learning procedures could overcome the difficulties of previous detection methods, and thus open up new possibilities for the practical application of sleep spindles as biomarkers in the psychiatric practice.

Advanced sleep spindle identification with neural networks.

Sleep spindles are neurophysiological phenomena that appear to be linked to memory formation and other functions of the central nervous system, and that can be observed in electroencephalographic recordings (EEG) during sleep. Manually identified spindle annotations in EEG recordings suffer from substantial intra- and inter-rater variability, even if raters have been highly trained, which reduces the reliability of spindle measures as a research and diagnostic tool. The Massive Online Data Annotation (MODA) project has recently addressed this problem by forming a consensus from multiple such rating experts, thus providing a corpus of spindle annotations of enhanced quality. Based on this dataset, we present a U-Net-type deep neural network model to automatically detect sleep spindles. Our model's performance exceeds that of the state-of-the-art detector and of most experts in the MODA dataset. We observed improved detection accuracy in subjects of all ages, including older individuals whose spindles are particularly challenging to detect reliably. Our results underline the potential of automated methods to do repetitive cumbersome tasks with super-human performance.

Automated scoring of pre-REM sleep in mice with deep learning.

Reliable automation of the labor-intensive manual task of scoring animal sleep can facilitate the analysis of long-term sleep studies. In recent years, deep-learning-based systems, which learn optimal features from the data, increased scoring accuracies for the classical sleep stages of Wake, REM, and Non-REM. Meanwhile, it has been recognized that the statistics of transitional stages such as pre-REM, found between Non-REM and REM, may hold additional insight into the physiology of sleep and are now under vivid investigation. We propose a classification system based on a simple neural network architecture that scores the classical stages as well as pre-REM sleep in mice. When restricted to the classical stages, the optimized network showed state-of-the-art classification performance with an out-of-sample F1 score of 0.95 in male C57BL/6J mice. When unrestricted, the network showed lower F1 scores on pre-REM (0.5) compared to the classical stages. The result is comparable to previous attempts to score transitional stages in other species such as transition sleep in rats or N1 sleep in humans. Nevertheless, we observed that the sequence of predictions including pre-REM typically transitioned from Non-REM to REM reflecting sleep dynamics observed by human scorers. Our findings provide further evidence for the difficulty of scoring transitional sleep stages, likely because such stages of sleep are under-represented in typical data sets or show large inter-scorer variability. We further provide our source code and an online platform to run predictions with our trained network.

Coupling and dynamics of cortical and autonomic signals are linked to central inhibition during the wake-sleep transition.

Maintaining temporal coordination across physiological systems is crucial at the wake-sleep transition. As shown in recent studies, the degree of coordination between brain and autonomic arousal influences attention, which highlights a previously unrecognised point of potential failure in the attention system. To investigate how cortical and autonomic dynamics are linked to the attentive process we analysed electroencephalogram, electrocardiogram and skin conductance data of 39 healthy adults recorded during a 2-h resting-state oddball experiment. We related cross-correlations to fluctuation periods of cortical and autonomic signals and correlated obtained measures to event-related potentials N1 and P2, reflecting excitatory and inhibitory processes. Increasing alignment of cortical and autonomic signals and longer periods of vigilance fluctuations corresponded to a larger and earlier P2; no such relations were found for N1. We compared two groups, with (I) and without measurable (II) delay in cortico-autonomic correlations. Individuals in Group II had more stable vigilance fluctuations, larger and earlier P2 and fell asleep more frequently than individuals in Group I. Our results support the hypothesis of a link between cortico-autonomic coupling and dynamics and central inhibition. Quantifying this link could help refine classification in psychiatric disorders with attention and sleep-related symptoms, particularly in ADHD, depression, and insomnia.

Optimal Extraction of Collective Oscillations from Unreliable Measurements.

We present a method that facilitates a phase description of collective, irregular-oscillatory dynamics from unreliable multichannel recordings. The collective oscillations may be represented in each channel with fluctuating amplitude, phase offsets, and substantial amounts of measurement noise. Our method performs well under such realistic conditions, as we exemplify with collective brain rhythms in multichannel electroencephalogram recordings.

Alpha-wave frequency characteristics in health and insomnia during sleep.

Appearances of alpha waves in the sleep electrencephalogram indicate physiological, brief states of awakening that lie in between wakefulness and sleep. These microstates may also cause the loss in sleep quality experienced by individuals suffering from insomnia. To distinguish such pathological awakenings from physiological ones, differences in alpha-wave characteristics between transient awakening and wakefulness observed before the onset of sleep were studied. In polysomnographic datasets of sleep-healthy participants (n = 18) and patients with insomnia (n = 10), alpha waves were extracted from the relaxed, wake state before sleep onset, wake after sleep-onset periods and arousals of sleep. In these, alpha frequency and variability were determined as the median and standard deviation of inverse peak-to-peak intervals. Before sleep onset, patients with insomnia showed a decreased alpha variability compared with healthy participants (P < 0.05). After sleep onset, both groups showed patterns of decreased alpha frequency that was lower for wake after sleep-onset periods of shorter duration. For patients with insomnia, alpha variability increased for short wake after sleep-onset periods. Major differences between the two groups were encountered during arousal. In particular, the alpha frequency in patients with insomnia rebounded to wake levels, while the frequency in healthy participants remained at the reduced level of short wake after sleep-onset periods. Reductions in alpha frequency during wake after sleep-onset periods may be related to the microstate between sleep and wakefulness that was described for such brief awakenings. Reduced alpha variability before sleep may indicate a dysfunction of the alpha generation mechanism in insomnia. Alpha characteristics may also prove valuable in the study of other sleep and attention disorders.

Robust design of polyrhythmic neural circuits.

Neural circuit motifs producing coexistent rhythmic patterns are treated as building blocks of multifunctional neuronal networks. We study the robustness of such a motif of inhibitory model neurons to reliably sustain bursting polyrhythms under random perturbations. Without noise, the exponential stability of each of the coexisting rhythms increases with strengthened synaptic coupling, thus indicating an increased robustness. Conversely, after adding noise we find that noise-induced rhythm switching intensifies if the coupling strength is increased beyond a critical value, indicating a decreased robustness. We analyze this stochastic arrhythmia and develop a generic description of its dynamic mechanism. Based on our mechanistic insight, we show how physiological parameters of neuronal dynamics and network coupling can be balanced to enhance rhythm robustness against noise. Our findings are applicable to a broad class of relaxation-oscillator networks, including Fitzhugh-Nagumo and other Hodgkin-Huxley-type networks.

Phase description of stochastic oscillations.

We introduce an invariant phase description of stochastic oscillations by generalizing the concept of standard isophases. The average isophases are constructed as sections in the state space, having a constant mean first return time. The approach allows us to obtain a global phase variable of noisy oscillations, even in the cases where the phase is ill defined in the deterministic limit. A simple numerical method for finding the isophases is illustrated for noise-induced switching between two coexisting limit cycles, and for noise-induced oscillation in an excitable system. We also discuss how to determine isophases of observed irregular oscillations, providing a basis for a refined phase description in data analysis.

Optimal phase description of chaotic oscillators.

We introduce an optimal phase description of chaotic oscillations by generalizing the concept of isochrones. On chaotic attractors possessing a general phase description, we define the optimal isophases as Poincaré surfaces showing return times as constant as possible. The dynamics of the resultant optimal phase is maximally decoupled from the amplitude dynamics and provides a proper description of the phase response of chaotic oscillations. The method is illustrated with the Rössler and Lorenz systems.

Effective phase dynamics of noise-induced oscillations in excitable systems.

We develop an effective description of noise-induced oscillations based on deterministic phase dynamics. The phase equation is constructed to exhibit correct frequency and distribution density of noise-induced oscillations. In the simplest one-dimensional case the effective phase equation is obtained analytically, whereas for more complex situations a simple method of data processing is suggested. As an application an effective coupling function is constructed that quantitatively describes periodically forced noise-induced oscillations.