Evaluation and Development of Integrated Mobile Technologies for Sleep and Cognition Sue Peters, Itamar Lerner, Iman Jashanmal, Mark A. Gluck Center for Molecular & Behavioral Neuroscience, Rutgers University, Newark, NJ, USA Introduction The importance of sleep to human health and well-being has long been recognized. Sleep, or lack there of, is a current focus in many human pathologies. However, there at least two central processes by which sleep affects the daily lives of healthy individuals: memory consolidation and emotional regulation. Over the last decade, there has been considerable progress in sleep research, but certain aspects of discovery have been limited by methodology. Polysomnography (PSG), the gold standard for sleep research, requires the subject to spend the night in a sleep lab wired to an EEG recorder, and thus limits the ecological validity and may only be practical for one to two nights. In order to significantly advance our current knowledge regarding the effects of long-term sleep patterns on cognition in a day-to-day resolution, new methodological tools are needed. In this study we will implement a Mobile Sleep Monitoring System and Cognitive. Assessment System, in order to address two important unanswered questions about the interactions between sleep, cognition, and mood. Question #1: What is the cumulative effect, over many days, of fluctuations in sleep patterns on cognitive and emotional well-being? Question #2: What are the effects of sleep on the gradual consolidation of memory in activities which require ongoing practice for days or weeks to master? See companion posters for additional information. Background In cognitive behavioral research, individuals may be asked about the prior night’s sleep through questionnaires or sleep-logs, prior to participating in a particular cognitive test. These methods lack objectivity and reliability. Aside from sleep-logs, the most commonly used method for measuring sleep over several days has been the use of limb actigraphy, which can sample a subject’s movement several times per second. This data is then interpreted to infer wake versus sleep states. In recent years several new “at-home” technologies have emerged, which provide objective, physiologic measures of sleep including 1 channel EEG devices, ECG, autonomic activity, and other movement based measurements, all of which incorporate algorithms for measuring various sleep characteristics. It is important to consider validation of these technologies against PSG, in light of the average 85% inter-rater reliability seen in PSG, which could be considered the upper-limit of an automated algorithm (Kelly et al. , 2012). Figure 1. On the left is a cartoon of a standard PSG setup in a sleep lab. On right are mobile devices which are used to measure physiologic activity during sleep. Clockwise: Actiwatch, Q-sensor, Zeo, Body. Media Fit. While individual cognitive assessment has traditionally been completed in research labs using desktop or laptop computers, these tasks are increasingly being made available for testing subjects athome or on mobile devices. Some challenges are the maintenance of integrity in the data collected, and consistency of task presentation. However, mobile devices may allow for the effective implementation of repeat testing with regular software updates, greater accessibility, data conservation, and better subject compliance. Technology Mobile Sleep Monitoring System This system consists of four components: Zeo Mobile Sleep Monitor, Actigraphy, Mobile Device and Device Management software. Zeo Mobile Sleep Monitor: collects a single-channel electrophysiological signal from the forehead, using electroencephalogram (EEG), eye movements and the frontalis muscle. The signal is transmitted wirelessly to a mobile phone (i. OS or Android), Kindle, or bedside base station. A reduced set of sleep stages are reported, based on a proprietary algorithm, and include wakefulness, REM sleep, light sleep, and deep or slow wave sleep (SWS). The device has been validated against 2 human PSG scorers with an 81% average agreement for an entire night, whereas the average agreement between the scorers was 83% (Shambroom et al. , 2011). Figure 2. A sample hypnogram of one night of sleep data which is provided to the user on the mobile device. Actigraphy and Electrodermal Activity (EDA): actigraphy provides an additional validation tool for sleep/wake distinction, allows for tracking of daytime naps, and provides a measure of daytime physical activity. Some actigraphs incorporate EDA measures (sympathetic nervous system), along with temperature flux. EDA “storms”, which occur during sleep, may correlate with certain types of daytime activity, including cognition and affect. This study may provide a unique opportunity to study this measure further as such exploration requires data collection over many nights, in order to understand individual patterns of activity, and their relevance to cognition and affect. . Cognitive Assessment System This system consists of a 8. 9 “ Kindle tablet (or other mobile device) and device management software. Mobile Device and Mobile Device Management: we chose the Kindle tablet running Android OS as our hardware and platform for our cognitive assessment applications. Some of these applications are covered in more detail in our companion posters. In order to participate, subjects were required to have a smartphone (i. OS or Android). Through our initial pilot, we found that some older devices were unreliable, but were able to run the sleep system on a Kindle and thus enabled non-smartphone users to participate in the study. The sleep and cognitive data is then transferred via wifi or wireless, from the tablet and phone to our secure data repository, on a daily basis, which minimizes data loss. We are evaluating existing systems for further mobile data and device management, which will allow for software update management, including daily stimulus updates for repeat testing, behavioral monitoring, Figure 3: . A participant using a Kindle to complete one of our and messaging features for cognitive tasks. compliance. . Preliminary Analysis Actigraphy provides an additional validation tool for sleep/wake distinction Immediate SWS (% of total from Zeo) correlates with improvement in cued recall no correlation experimental group, r=-. 63, n=11, p<. 03 experimental group, r=+. 55, n=11, p<. 07 wake sleep Figure 4. Four different methods were used for analysis of the actigraphy data. We found the Cole-Kripke, zero-crossing scores, to be the most robust in terms of differentiating wake and sleep when compared to the simultaneous collection of Zeo (hypnogram) data. In this data sample from one subject, the steep decline in motion can be seen at the 15 minute mark, and the algorithm accurately codes this as sleep. This delineation is more precise than the time of sleep onset noted by Zeo. recognition of sleep-onset Figure 5. In order to test the effectiveness of our Mobile Sleep Monitoring and Cognitive Assessment System, we ran a short term study measuring sleep physiology over 4 -5 days per subject, with the final day incorporating a word-pair retention task. Our preliminary results show memory retention scores correlated with slow wave sleep, similar to the classic sleep and paired associates studies done with PSG. (Plihal & Born, 1997). Conclusions • Using the Zeo mobile sleep monitor, we were able to replicate findings from PSG studies which show a correlation between word pair retention and slow wave sleep (see companion poster). • The actigraphy measurement does provide us with a more reliable measure of transition between wake and sleep, and may augment the Zeo sleep hardware and software, in terms of accuracy of total sleep time. • The electrodermal activity monitor may provide additional physiologic sleep data, relevant to sleep and emotion. This requires further testing and analysis. • We have ported 4 cognitive tasks to the Android platform, and have found the Android to be a reliable OS for our cognitive tasks, and the 8. 9” Kindle Fire to be a reliable and affordable device for this project. References Kelly, J. , Strecker, R. E. , Bianchi, M. (2012). Recent Developments in Home Sleep-Monitoring Devices. ISRN Neurology. Shambroom, J. R. , Fabregas, S. E. , & Johnstone, J. (2011). Validation of an automated wireless system to monitor sleep in healthy adults. J Sleep Res. Plihal, W & Born, J. , (1997). Effects of Early and Late Nocturnal Sleep on Declarative and Procedural Memory. Journal of Cognitive Neuroscience. Acknowledgements Supported by Grant #7367437 for “Long-term Mobile Monitoring and Analysis of Sleep-Cognition Relationship” from the National Science Foundation's Smart Health and Wellbeing program to M. A. G. Contact Sue Peters, Graduate Student, sp@suepeters. com Mark Gluck, PI, gluck@pavlov. rutgers. edu
Скачать презентацию Evaluation and Development of Integrated Mobile Technologies for
8a332d6141f48782b65dc9d4fd79a8f0.ppt