Sleep staging is to score the sleep state of a subject into different sleep stages such as Wake and Rapid Eye Movement (REM). It plays an indispensable role in the diagnosis and treatment of sleep diseases. As manual sleep staging through well-trained sleep experts is time consuming, tedious, and subjective, many automatic methods have been developed for accurate, efficient, and objective sleep staging. Recently, deep learning based methods have been successfully proposed for EEG based sleep staging with promising results. However, most of these methods directly take EEG raw signals as input of convolutional neural networks (CNNs) without considering the domain knowledge of EEG staging. Apart from that, to capture temporal information, most of the existing methods utilize recurrent neural networks such as LSTM (Long Short Term Memory) which are not effective for modelling global temporal context and difficult to train. Therefore, inspired by the clinical guidelines of sleep staging such as AASM (American Academy of Sleep Medicine) rules where different stages are generally characterized by EEG waveforms of various frequencies, we propose a multi-scale deep architecture by decomposing an EEG signal into different frequency bands as input to CNNs. To model global temporal context, we utilize the multi-head self-attention module of the transformer model to not only improve performance, but also shorten the training time. In addition, we choose residual based architecture which makes training end-to-end. Experimental results on two widely used sleep staging datasets, Montreal Archive of Sleep Studies (MASS) and sleep-EDF datasets, demonstrate the effectiveness and significant efficiency (up to 12 times less training time) of our proposed method over the state-of-the-art.

Introduction: The use of simulation in health professional education has increased rapidly over the past 2 decades. While simulation has predominantly been used to train health professionals and students for a variety of clinically related situations, there is an increasing trend to use simulation as an assessment tool, especially for the development of technical-based skills required during clinical practice. However, there is a lack of evidence about the effectiveness of using simulation for the assessment of competency. Therefore, the aim of this systematic review was to examine simulation as an assessment tool of technical skills across health professional education. Methods: A systematic review of Cumulative Index to Nursing and Allied Health Literature (CINAHL), Education Resources Information Center (ERIC), Medical Literature Analysis and Retrieval System Online (Medline), and Web of Science databases was used to identify research studies published in English between 2000 and 2015 reporting on measures of validity, reliability, or feasibility of simulation as an assessment tool. The McMasters Critical Review for quantitative studies was used to determine methodological value on all full-text reviewed articles. Simulation techniques using human patient simulators, standardized patients, task trainers, and virtual reality were included. Results: A total of 1,064 articles were identified using search criteria, and 67 full-text articles were screened for eligibility. Twenty-one articles were included in the final review. The findings indicated that simulation was more robust when used as an assessment in combination with other assessment tools and when more than one simulation scenario was used. Limitations of the research papers included small participant numbers, poor methodological quality, and predominance of studies from medicine, which preclude any definite conclusions. Conclusion: Simulation has now been embedded across a range of health professional education and it appears that simulation-based assessments can be used effectively. However, the effectiveness as a stand-alone assessment tool requires further research.

Background and Purpose: The exact prevalence of sleep disorders following stroke or transient ischemic attack (TIA) remains unclear. We aimed to determine the prevalence of sleep-disordered breathing, insomnia, periodic leg movement during sleep, and restless leg syndrome following stroke or TIA in acute, subacute, and chronic phases and examine the moderating effects of patient characteristics (eg, age) and methodological features (eg, study quality) on the prevalence. Methods: We performed a systematic review and meta-analysis. Embase and PubMed were searched from inception to December 18, 2019. We included 64 047 adults in 169 studies (prospective, retrospective, case-control, and cross-sectional study designs) reporting the prevalence of sleep disorders following stroke or TIA. Results: In the acute phase, the overall prevalence of mild, moderate, and severe sleep-disordered breathing was 66.8%, 50.3%, and 31.6% (95% CIs, 63.8–69.7, 41.9–58.7, and 24.9–39.1). In the subacute phase, the prevalence of mild, moderate, and severe sleep-disordered breathing was 65.5%, 44.3%, and 36.1% (95% CIs, 58.9–71.5, 36.1–52.8, and 22.2–52.8). In the chronic phase, the summary prevalence of mild, moderate, and severe sleep-disordered breathing was 66.2%, 33.1%, and 25.1% (95% CIs, 58.6–73.1, 24.8–42.6, and 10.9–47.6). The prevalence rates of insomnia in the acute, subacute, and chronic phases were 40.7%, 42.6%, and 35.9% (95% CIs, 31.8–50.3, 31.7–54.1, and 28.6–44.0). The pooled prevalence of periodic leg movement during sleep in the acute, subacute, and chronic phases was 32.0%, 27.3%, and 48.2% (95% CIs, 7.4–73.5, 11.6–51.7, and 33.1–63.5). The summary prevalence of restless leg syndrome in the acute and chronic phases was 10.4% and 13.7% (95 CIs, 6.4–16.4 and 2.3–51.8). Age, sex, comorbidities, smoking history, and study region had significant moderating effects on the prevalence of sleep disorders. Conclusions: Sleep disorders following stroke or TIA are highly prevalent over time. Our findings indicate the importance of early screening and treating sleep disorders following stroke or TIA.