In this study, we developed the Bern continuous and high-resolution wake-sleep (BERN) scoring criteria as a first step toward closing an important gap of the scoring criteria established by the American Academy of Sleep Medicine, particularly required for classifying the wake-sleep transition zone. We were able to objectify our visual scoring criteria by comparison with a quantitative analysis of the electroencephalography. The application of the BERN scoring criteria led to new and promising insights into the borderland between wakefulness and sleep. A further refinement combined with automatic detection and inclusion of behavior and performance measures could offer a powerful new tool for clinical sleep medicine and research, improving the differential diagnosis and the assessment of treatment and fitness to drive.
Preceding the detection of rapid eye movement (REM) sleep by Aserinsky and Kleitmann, Loomis et al. proposed a sleep classification, consisting of wakefulness and four different sleep stages. The stage B1 was related to “drowsiness” and the stage B2 to “sleep onset”. Compared with later sleep scoring classifications, this initial classification of the wake-sleep transition zone contained more details. In 1968, Rechtschaffen and Kales (R&K) established the first globally applied criteria for classifying the electroencephalography (EEG) into wakefulness, non-REM sleep stages (N1–4), and REM sleep. It was only in 2007, after almost 40 years, that the American Academy of Sleep Medicine (AASM) revised the criteria of R&K. Since then, the AASM has generally revised their criteria on a yearly basis with the latest revision (version 2.5) being published in April 2018. The AASM scoring criteria, and formerly the R&K criteria, represent the international standard for scoring sleep and wakefulness in the most important diagnostic sleep-wake tests, such as polysomnography (PSG), the Multiple Sleep Latency Test (MSLT), and the Maintenance of Wakefulness Test (MWT). In addition to the clinical use, the scoring criteria are widely used in clinical and basic sleep research.
Historically, EEGs were recorded on paper with a standard running speed of 10 mm/s, mostly with an epoch length being defined as 30 s, equivalent to a one-page EEG recording. Surprisingly, the defined length of 30 s has outlasted the digitalization of EEG recordings and variations in computer screen size. Similarly, the clinical scoring criteria underwent few changes and still do not adequately address the wake-sleep transition zone, despite the technical advances and expanded knowledge on the process of falling asleep. In addition to the absence of an intermediate stage between wakefulness and N1, the low temporal resolution (30 s) does not take the rapid fluctuations between different stages into consideration. This is of particular importance in the MWT, during which individuals are instructed to stay awake despite their potentially excessive sleepiness, thus prolonging the time spent within the wake-sleep transition zone. Although limited, normative data are available for scoring epochs shorter than 30 s in the MWT, that is, a minimum of 10 s resolution. The rapid fluctuations between wakefulness and sleep characterize the typical instability in the wake-sleep transition zone and may result in so-called microsleeps or microsleep episodes (MSEs). Wake-sleep fluctuations can persist over protracted time periods, particularly during daily performance tasks, such as driving, and increase the chance for MSEs to occur, potentially resulting in fatal accidents. In contrast to the MWT and performance tasks, a rather rapid transition from wakefulness to deeper sleep stages can be observed in PSG or the MSLT, during which individuals are allowed to fall sleep.
Unfortunately, there is no generally accepted definition regarding the true meaning of an MSE. The definition depends mainly on the type of the recorded signal and can be classified into the following three categories: (1) EEG or other neurophysiological parameters (visual and automatic), (2) eye, eyelid, or face/body behavior, and (3) psychomotor performance measures. Several parameters from any type of recording can be utilized alone or in combination.
In clinical sleep medicine, MSEs are predominantly defined by short-lasting EEG patterns resembling sleep (mainly N1). However, it is important to be aware that signs and symptoms, as well as performance lapses can occur before the appearance of a clear-cut sleep-like EEG pattern in the superficial EEG. In our understanding, sleepiness is a condition that precedes sleep and reflects increased sleep pressure, which, depending on the task and the level of sleepiness, cannot always be perceived subjectively. While sleepiness and drowsiness are often used synonymously, in our view, the latter represents a variant of sleepiness during which MSEs are not necessarily visible in the EEG. Drowsiness is frequently accompanied by physiological measures such as eyelid drooping, rolling eye movements, slowing in heart rate or respiration, and performance impairments. The discrepancy between the observations of behavioral changes in the absence of a correlating pattern in the surface EEG can be explained by local sleep starting in subcortical brain structures, which is not visible in the surface EEG before spreading to the cortex. However, higher levels of sleepiness result in EEG patterns, which can also be objectified in superficial EEG while behavioral changes are simultaneously observable. Accordingly, a differentiation of “EEG-defined MSEs,” “behaviour-defined MSEs,” and “performance-defined MSEs” is primarily dependent on the assessment method but does not necessarily reflect the underlying physiological processes. In clinical sleep medicine and research, the EEG still represents the method of choice for defining MSEs. Behavioral or performance measurement tools are often not available. However, combining the EEG with face videography can improve the reliable identification of MSEs in the MWT. Therefore, it has long been standard practice to record both during the MWT in the Sleep-Wake-Epilepsy-Centre of the Bern University Hospital.
The present study focused on the wake-sleep transition zone, which ranges from full wakefulness through the first signs of sleepiness to severe sleepiness and sleep, the latter defined according to the AASM criteria. Among all tests available in a clinical environment, the MWT best resembles a passive real-life condition (e.g. surveillance task), in which MSEs are of particular relevance. Furthermore, the sleep latency in the MWT not only correlates with driving performance, but this correlation is stronger than the one of sleep latency in the MSLT and driving performance, although this is without taking MSEs into account. Therefore, we aimed to identify MSEs with a high temporal resolution and great specificity in the MWT.
The primary aim of this study was to define practicable visual EEG scoring criteria for the wake-sleep transition zone, taking a first step toward the closure of the gap in the AASM scoring criteria. The first objective was to formulate the Bern continuous and high-resolution wake-sleep (BERN) criteria for visually scoring MSEs and similar less clearly defined EEG patterns in the MWT with a high temporal resolution (i.e. minimal duration of 1 s). The second objective was to compare the visual scoring of MSEs according to the newly developed BERN criteria with quantitative EEG and electrooculography (EOG) analyses.
The secondary aim of this study was to further analyze and characterize the borderland between wakefulness and sleep. The first objective of the secondary aim was to descriptively analyze MSEs and investigate their impact on the “sleep latency” by comparison of the latency to the first MSE with the latency to sleep. The second objective was to analyze the temporal distribution and dynamics of MSEs.
You can view the study methods and full article at Oxford Sleep here.