Abstract Introduction An estimated 34-million U.S. adults suffer from chronic insomnia. The first-line treatment for this disease is Cognitive Behavioral Therapy for Insomnia (CBT-I). However, poor access, high cost, and low adherence prevent its widespread use. Full Sleep is a novel Internet of Things (IoT) device that includes an intelligence-sensing radar to passively track sleep patterns and deliver in the moment behavioral prompting to promote adherence to CBT-I directives. Methods The effect of the IoT device was examined on participants with moderate to severe insomnia symptoms (n = 65). Participants were recruited through a website and consented to a 6-8 week program. The IoT device was hypothesized to improve the Insomnia Severity Index (ISI) (primary outcome), evaluated at baseline and post-treatment. The IoT device was also hypothesized to improve consensus sleep diary metrics. Exploratory hypotheses included increasing total sleep time and reducing depression and anxiety on survey assessments. Results A within-groups paired-comparison t-test showed significant improvements in the ISI from baseline 19.3 (SD = 2.5) to post-treatment 7.4 (SD = 4.6), t(43) = 16.89, p < .001, d = 2.5, a 62% decrease. A clinically significant change is 6-points, demonstrating a dramatic improvement, and a larger effect than reported by digital therapeutics like Somryst, which found a 45% decrease in the ISI. There was also evidence of a causal relationship between improvements in the ISI and the use of the IoT device, where the more that participants engaged with the feature that promoted getting out of bed during restless nights, the greater the improvement in the ISI score r(42) = .37, p < .05. Importantly, the intervention also increased total sleep time, from 6.3 hours (SD = 1.4) to 6.8 hours (SD = 1.4), t(51) = 4.14, p < .05, and sleep efficiency from 73.6% (SD = 13.7%) to 88.6% (SD = 8.9%), t(51) = 9.9, p < .05. Participants had a reduction in dysfunctional beliefs about sleep, depressive symptoms, and anxious symptoms (ps < .05). Conclusion These results support that innovative technologies that work alongside CBT-I directives can reduce insomnia severity and increase total sleep time. Support (if any) This research was supported by Koko Home Inc.

Abstract Introduction Of 34M US adults affected by insomnia, 75% are older adults. Cognitive Behavioral Therapy for Insomnia (CBTi) is recommended because polypharmacy and fall risks accompany pharmacotherapies. We evaluated telehealth CBTi with an interactive patient-therapist application, SleepSpace, which integrates data from wearables and Internet of Things (IoT) devices. Methods This RCT (NCT05015803) followed community-dwelling participants 60-90 years old with an Insomnia Severity Index (ISI) score ≥11. Absence of mild cognitive impairment was affirmed with the Montreal Cognitive Assessment (MoCA) Blind v.8 (score ≥ 18). Participants wore actigraphy and an Apple Watch throughout and independently completed a weekly electronic ISI. They attended 7 weekly, ~1hr video-conference sessions (1 intake, 6 procedural) with a clinical therapist. Participants were randomly assigned to one of 3 study conditions (age-, gender-stratified): 1) education about sleep hygiene only (20%; “Hygiene”), 2) telehealth CBTi (40%; “CBTi”), and 3) telehealth CBTi with phone/IoT platform application enhancement (40%; “CBTi+”) including meditations, sound machines, smart light bulbs, an electronic diary, with visualizations, metrics, and wearable data shared with participants in the CBTi+ condition. Linear mixed models compared ISI change across time by group. Results Of 60 individuals enrolled, 54 were randomized and retained (39F, mean±SD age=71±4y). ISI slopes for both CBTi (-.09/day) and CBTi+ (-.09/day) declined at a significantly steeper rate than Hygiene (-.05/day; each p<.05), but did not differ from one another. Significantly more CBTi+ participants exhibited full remission (ISI < 8; 18/21, 85.7%) than in the Hygiene group (5/11, 45.4%; p=.03 Fisher’s Exact); CBTi alone (16/22, 72.7%) did not significantly differ from Hygiene, although with limited statistical power. Diary-reported sleep measures to calculate self-reported sleep efficiency (sSE) in the final week at end of treatment revealed differences in mean±SD for Hygiene (81±05%) vs. CBTI (88±07%), and vs. CBTI+ (90±04%, p< 0.05, t-test). Conclusion This research supports the efficacy of a remote, technology-assisted telehealth CBTi platform to improve insomnia symptoms comparable to standard telehealth-CBTi in older adults with insomnia. The platform provides enhanced data access for therapists and opportunities for data-driven engagement with patients. Support (if any) R44 AG056250, UL1TR002014

Aging populations are at increased risk of sleep deficiencies (e.g., insomnia) that are associated with a variety of chronic health risks, including Alzheimer's disease and related dementias (ADRD). Insomnia medications carry additional risk, including increased drowsiness and falls, as well as polypharmacy risks. The recommended first line treatment for insomnia is cognitive behavioral therapy for insomnia (CBTi), but access is limited. Telehealth is one way to increase access, particularly for older adults, but to date telehealth has been typically limited to simple videoconferencing portals. While these portals have been shown to be non-inferior to in-person treatment, it is plausible that telehealth could be significantly improved. This work describes a protocol designed to evaluate whether a clinician-patient dashboard inclusive of several user-friendly features (e.g., patterns of sleep data from ambulatory devices, guided relaxation resources, and reminders to complete in-home CBTi practice) could improve CBTi outcomes for middle-to older-aged adults (N = 100). Participants were randomly assigned to one of three telehealth interventions delivered through 6-weekly sessions: (1) CBTi augmented with a clinician patient dashboard, smartphone application, and integrated smart devices; (2) standard CBTi (i.e., active comparator); or (3) sleep hygiene education (i.e., active control). All participants were assessed at screening, pre study evaluation, baseline, throughout treatment, and at 1-week post-treatment. The primary outcome is the Insomnia Severity Index. Secondary and exploratory outcomes span sleep diary, actiwatch and Apple watch assessed sleep parameters (e.g., efficiency, duration, timing, variability), psychosocial correlates (e.g., fatigue, depression, stress), cognitive performance, treatment adherence, and neurodegenerative and systemic inflammatory biomarkers.

GOAL AND AIMS:Commonly used actigraphy algorithms are designed to operate within a known in-bed interval. However, in free-living scenarios this interval is often unknown. We trained and evaluated a sleep/wake classifier that operates on actigraphy over ∼24-hour intervals, without knowledge of in-bed timing. FOCUS TECHNOLOGY:Actigraphy counts from ActiWatch Spectrum devices. REFERENCE TECHNOLOGY:Sleep staging derived from polysomnography, supplemented by observation of wakefulness outside of the staged interval. Classifications from the Oakley actigraphy algorithm were additionally used as performance reference. SAMPLE:Adults, sleeping in either a home or laboratory environment. DESIGN:Machine learning was used to train and evaluate a sleep/wake classifier in a supervised learning paradigm. The classifier is a temporal convolutional network, a form of deep neural network. CORE ANALYTICS:Performance was evaluated across ∼24 hours, and additionally restricted to only in-bed intervals, both in terms of epoch-by-epoch performance, and the discrepancy of summary statistics within the intervals. ADDITIONAL ANALYTICS AND EXPLORATORY ANALYSES:Performance of the trained model applied to the Multi-Ethnic Study of Atherosclerosis dataset. CORE OUTCOMES:Over ∼24 hours, the temporal convolutional network classifier produced the same or better performance as the Oakley classifier on all measures tested. When restricting analysis to the in-bed interval, the temporal convolutional network remained favorable on several metrics. IMPORTANT SUPPLEMENTAL OUTCOMES:Performance decreased on the Multi-Ethnic Study of Atherosclerosis dataset, especially when restricting analysis to the in-bed interval. CORE CONCLUSION:A classifier using data labeled over ∼24-hour intervals allows for the continuous classification of sleep/wake without knowledge of in-bed intervals. Further development should focus on improving generalization performance.

Abstract Introduction Wrist-worn research-grade actigraphy devices are commonly used to identify sleep and wakefulness in freely-living people. However, common existing algorithms were developed primarily to classify sleep-wake within a defined in-bed period with PSG, and exhibit relatively high sensitivity (accuracy on sleep epochs) but relatively low specificity (accuracy on wake epochs). This classification imbalance results in the algorithms performing poorly when attempting to classify data that does not have a predefined sleep period, such as over a 24-hour interval. Here, we develop a 24-hour actigraphy classifier to overcome limitations in specificity (accuracy on wake epochs), which typically afflict in-bed focused algorithms. Methods Four datasets scored via either PSG or direct observation of daytime wakefulness were combined (n=52 participants of mean age 49.8yrs, age range 19 - 86; 52% male; 221 total days/nights). Actigraphy (counts) and PSG (RPSGT-staged epochs) were temporally aligned. A model was trained to transform a time-series actigraphy counts to a time series of sleep-wake classifications, using the TensorFlow library for Python. 5-fold cross-validation was used to train and evaluate the model. Classification performance was compared to the output of the Spectrum device (Philips-Respironics) using the Oakley algorithm with a wake threshold of ‘medium’. Results The developed classifier was compared to the Spectrum classifications across the 24-hour intervals. The developed classifier had higher accuracy (95.4% vs. 76.8%), higher specificity (95.9% vs. 68.9%) and higher balanced-accuracy (95.2% vs. 81.6%) relative to the Spectrum classifications, each assessed via paired-sample t-test. Sensitivity did not statistically differ (94.5% vs. 94.4%). Conclusion The model trained and evaluated on 24-hour data outperformed the existing algorithm output in terms of overall accuracy, specificity, and balanced accuracy, while sensitivity did not significantly differ. A model trained on 24-hour data may be more appropriate for analyses of freely living people, or older populations where napping is more common. Developing an accurate 24-hour sleep/wake classifier fosters new opportunities to evaluate sleep patterns in the absence of self-reports or assumptions about time in bed. Support (If Any) UL1TR002014, NSF#1622766, R43/44-AG056250

STUDY OBJECTIVES:Multisensor wearable consumer devices allowing the collection of multiple data sources, such as heart rate and motion, for the evaluation of sleep in the home environment, are increasingly ubiquitous. However, the validity of such devices for sleep assessment has not been directly compared to alternatives such as wrist actigraphy or polysomnography (PSG).METHODS:Eight participants each completed four nights in a sleep laboratory, equipped with PSG and several wearable devices. Registered polysomnographic technologist-scored PSG served as ground truth for sleep-wake state. Wearable devices providing sleep-wake classification data were compared to PSG at both an epoch-by-epoch and night level. Data from multisensor wearables (Apple Watch and Oura Ring) were compared to data available from electrocardiography and a triaxial wrist actigraph to evaluate the quality and utility of heart rate and motion data. Machine learning methods were used to train and test sleep-wake classifiers, using data from consumer wearables. The quality of classifications derived from devices was compared.RESULTS:For epoch-by-epoch sleep-wake performance, research devices ranged in d' between 1.771 and 1.874, with sensitivity between 0.912 and 0.982, and specificity between 0.366 and 0.647. Data from multisensor wearables were strongly correlated at an epoch-by-epoch level with reference data sources. Classifiers developed from the multisensor wearable data ranged in d' between 1.827 and 2.347, with sensitivity between 0.883 and 0.977, and specificity between 0.407 and 0.821.CONCLUSIONS:Data from multisensor consumer wearables are strongly correlated with reference devices at the epoch level and can be used to develop epoch-by-epoch models of sleep-wake rivaling existing research devices.

Purpose: In non-rapid eye movement (NREM) stage 3 sleep (N3), phase-locked pink noise auditory stimulation can amplify slow oscillatory activity (0.5-1 Hz). Open-loop pink noise auditory stimulation can amplify slow oscillatory and delta frequency activity (0.5-4 Hz). We assessed the ability of pink noise and other sounds to elicit delta power, slow oscillatory power, and N3 sleep. Participants and Methods: Participants (n = 8) underwent four consecutive inpatient nights in a within-participants design, starting with a habituation night. A registered polysomnographic technologist live-scored sleep stage and administered stimuli on randomized counterbalanced Enhancing and Disruptive nights, with a preceding Habituation night (night 1) and an intervening Sham night (night 3). A variety of non-phase-locked pink noise stimuli were used on Enhancing night during NREM; on Disruptive night, environmental sounds were used throughout sleep to induce frequent auditory-evoked arousals. Results: Total sleep time did not differ between conditions. Percentage of N3 was higher in the Enhancing condition, and lower in the Disruptive condition, versus Sham. Standard 0.8 Hz pink noise elicited low-frequency power more effectively than other pink noise, but was not the most effective stimulus. Both pink noise on the "Enhancing" night and sounds intended to Disrupt sleep administered on the "Disruptive" night increased momentary delta and slow-wave activity (ie, during stimulation versus the immediate pre-stimulation period). Disruptive auditory stimulation degraded sleep with frequent arousals and increased next-day vigilance lapses versus Sham despite preserved sleep duration and momentary increases in delta and slow-wave activity. Conclusion: These findings emphasize sound features of interest in ecologically valid, translational auditory intervention to increase restorative sleep. Preserving sleep continuity should be a primary consideration if auditory stimulation is used to enhance slow-wave activity.

The vigilance decrement in sustained attention tasks is a prevalent example of cognitive fatigue in the literature. A critical challenge for current theories is to account for differences in the magnitude of the vigilance decrement across tasks that involve memory (successive tasks) and those that do not (simultaneous tasks). The empirical results described in this paper examine this issue by comparing performance, including eye movement data, between successive and simultaneous tasks that require multiple fixations to encode the stimulus for each trial. The findings show that differences in the magnitude of the vigilance decrement between successive and simultaneous tasks were observed only when a response deadline was imposed in the analysis of reaction times. This suggests that memory requirements did not exacerbate the deleterious impacts of time on task on the ability to accurately identify the critical stimuli. At the same time, eye tracking data collected during the study provided evidence for disruptions in cognitive processing that manifested as increased delays between fixations on stimulus elements and between encoding the second stimulus element and responding. These delays were particularly pronounced in later stages of encoding and responding. The similarity of the findings for both tasks suggests that the vigilance decrement may arise from common mechanisms in both cases. Differences in the magnitude of the decrement arise as a function of how degraded cognitive processing interacts with differences in the information processing requirements and other task characteristics. The findings are consistent with recent accounts of the vigilance decrement, which integrate features of prior theoretical perspectives.

Despite its high sensitivity and validity in the context of sleep loss, the Psychomotor Vigilance Test (PVT) could be improved. The aim of the present study was to validate a new smartphone PVT-type application called sleep-2-Peak (s2P) by determining its ability to assess fatigue-related changes in alertness in a context of extended wakefulness. Short 3-min versions of s2P and of the classic PVT were administered at every even hour during a 35-h total sleep deprivation protocol. In addition, subjective measures of sleepiness were collected. The outcomes on these tests were then compared using Pearson product-moment correlations, t tests, and repeated measures within-groups analyses of variance. The results showed that both tests significantly correlated on all outcome variables, that both significantly distinguished between the alert and sleepy states in the same individual, and that both varied similarly through the sleep deprivation protocol as sleep loss accumulated. All outcome variables on both tests also correlated significantly with the subjective measures of sleepiness. These results suggest that a 3-min version of s2P is a valid tool for differentiating alert from sleepy states and is as sensitive as the PVT for tracking fatigue-related changes during extended wakefulness and sleep loss. Unlike the PVT, s2P does not provide feedback to subjects on each trial. We discuss how this feature of s2P raises the possibility that the performance results measured by s2P could be less impacted by motivational confounds, giving this tool added value in particular clinical and/or research settings.

We describe a vigilance experiment of a successive task and a simultaneous task. Successive tasks require comparing the current stimulus on the screen to a representation in memory (i.e. making a declarative memory retrieval), whereas simultaneous tasks require making a comparative judgment based on information that is available on the screen. When analyzing the data from this experiment using conventional methods, there was an effect of time-on-task (i.e. block), an effect of task type, and an interaction between block and task type. These findings were consistent with previously reported studies regarding the successive and simultaneous vigilance task distinction, which interpret such findings as evidence that the decrement is more severe for successive tasks. But different results and conclusions are made when more appropriate analyses of the data are used, such as: including block as an interval variable instead of a categorical variable and making the dependent variable detection of critical signals instead of using A’. When these analysis techniques were used, there was no effect of task type and there was no interaction with time on task. This raises questions about many of the findings in the literature, especially those regarding the successive and simultaneous distinction.

Crandall et al. and Cummings & Mitchell introduced fan-out as a measure of the maximum number of robots a single human operator can supervise in a given single-human-multiple-obot system. Fan-out is based on the time constraints imposed by limitations of the robots and of the supervisor, e. g., limitations in attention. Adapting their work, we introduced a dynamic model of operator overload that predicts failures in supervisory control in real time, based on fluctuations in time constraints and in the supervisor's allocation of attention, as assessed by eye fixations. Operator overload was assessed by damage incurred by unmanned aerial vehicles when they traversed hazard areas. The model generalized well to variants of the baseline task. We then incorporated the model into the system where it predicted in real time, when an operatorwould fail to prevent vehicle damage and alerted the operator to the threat at those times. These model-based adaptive cues reduced the damage rate by one-half relative to a control condition with no cues.

Performance on tasks that require sustained attention can be impacted by various factors that include: signal duration, the use of declarative memory in the task, the frequency of critical stimuli that require a response, and the event-rate of the stimuli. A viable model of the ability to maintain vigilance ought to account for these phenomena. In this paper, we focus on one of these critical factors: signal duration. For this we use results from Baker (1963), who manipulated signal duration in a clock task where the second hand moved in a continuous swipe motion. The critical stimuli were stoppages of the hand that lasted for 200, 300, 400, 600, or 800 ms. The results provided evidence for an interaction between condition and time-on-task, where performance declined at a faster rate as the signal duration decreased. In this paper, we describe an ACT-R model that uses fatigue mechanisms from Gunzelmann et al. (2009) that were proposed to account for the impact of sleep loss on sustained attention performance. The research demonstrates how those same mechanisms can be used to understand vigilance task performance. This illustrates an important foundation for predicting and tracking vigilance decrements in applied settings, and validates a mechanism that creates a theoretical link between the vigilance decrement and sleep loss.

Abstract : Crandall et al. and Cummings & Mitchell introduced fan-out as a measure of the maximum number of robots a single human operator can supervise in a given single-human-multiple-robot system, based on the time constraints imposed by limitations of the robots and of the supervisor, e.g., limitations in attention. Adapting their work, we introduced a dynamic model of operator overload that predicts failures in supervisory control in real time, based on fluctuations in time constraints and in the supervisor's allocation of attention, assessed by eye fixations. Operator overload was assessed by damage incurred by vehicles when they traversed hazard areas. The model generalized well to different tasks. We then incorporated the model into the system where it predicted in real-time when an operator would fail to prevent vehicle damage and alerted the operator to the threat at those times. These model-based adaptive cues reduced the damage rate by one half relative to a control condition.

Objective: We describe a novel concept, situation awareness recovery (SAR), and we identify perceptual and cognitive processes that characterize SAR.Background: Situation awareness (SA) is typically described in terms of perceiving relevant elements of the environment, comprehending how those elements are integrated into a meaningful whole, and projecting that meaning into the future. Yet SA fluctuates during the time course of a task, making it important to understand the process by which SA is recovered after it is degraded.Method: We investigated SAR using different types of interruptions to degrade SA. In Experiment 1, participants watched short videos of an operator performing a supervisory control task, and then the participants were either interrupted or not interrupted, after which SA was assessed using a questionnaire. In Experiment 2, participants performed a supervisory control task in which they guided vehicles to their respective targets and either experienced an interruption, during which they performed a visual search task in a different panel, or were not interrupted.Results: The SAR processes we identified included shorter fixation durations, increased number of objects scanned, longer resumption lags, and a greater likelihood of refixating on objects that were previously looked at.Conclusions: We interpret these findings in terms of the memory-for-goals model, which suggests that SAR consists of increased scanning in order to compensate for decay, and previously viewed cues act as associative primes that reactivate memory traces of goals and plans.

We describe a mobile health application that collects data relevant to the treatment of insomnia and other sleep-related problems. The application is based on the principles from neuroergonomics, which emphasizes assessment of the brain’s alertness system in everyday, naturalistic environments, and ubiquitous computing. Application benefits include the ability to incorporate both embedded data collection and retrospective manual data input—thus providing the user with a rewarding data access process. The retrospective data input feature was evaluated by comparing an older version of the retrospective editing interface with a newly developed one. The time course of user interactions was precisely measured by exporting time stamps of user interactions using the Google App Engine. We also developed models that closely fit the time course of user interactions using the Goals, Operators, Methods, and Selection rules (GOMS) modeling method. The user data and GOMS models demonstrated that the retrospective sleep tracking feature of the new interface was faster to use but that the retrospective habit tracking feature was slower. Survey results indicated that participants enjoyed using the newly developed interface more than the old interface for the assessment of both sleep and habits. These findings indicate that a mobile application should be designed not only to reduce the amount of time it takes a user to input data, but also to conform to the user’s mental models of its behavior.

http://hfs.sagepub.com/content/early/2014/06/17/0018720814539505 The online version of this article can be found at: DOI: 10.1177/0018720814539505 June 2014 published online 18 Human Factors: The Journal of the Human Factors and Ergonomics Society Gerald Matthews, Lauren E. Reinerman-Jones, Daniel J. Barber and Julian Abich IV Divergent The Psychometrics of Mental Workload: Multiple Measures Are Sensitive but

: In daily conversations, what information do people use to assess their conversational partner's explanations? We explore how a metacognitive cue, in particular the partner's confidence or uncertainty, can modulate the credibility of an explanation. Two experiments showed that explanations are accepted more often when delivered by an uncertain conversational partner. Participants in Experiment 1 demonstrated the general effect by interacting with a pseudoautonomous robotic confederate. Experiment 2 used the same methodology to show that the effect was applicable to explanatory reasoning and not other sorts of inferences. Results are consistent with an account in which reasoners use relative confidence as a metacognitive cue to infer their conversational partner's depth of processing.

Adaptive automation (AA) can improve performance while addressing the problems associated with a fully automated system. The best way to invoke AA is unclear, but two ways include critical events and the operator's state. A hybrid model of AA invocation, the dynamic model of operator overload (DMOO), that takes into account critical events and the operator's state was recently shown to improve performance. The DMOO initiates AA using critical events and attention allocation, informed by eye movements. We compared the DMOO with an inaccurate automation invocation system and a system that invoked AA based only on critical events. Fewer errors were made with DMOO than with the inaccurate system. In the critical event condition, where automation was invoked at an earlier point in time, there were more memory and planning errors, while for the DMOO condition, which invocated automation at a later point in time, there were more perceptual errors. These findings provide a framework for reducing specific types of errors through different automation invocation.