A balanced sleep/wakefulness classification method based on actigraphic data in adolescents

Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach

BACKGROUND

The current gold standard for measuring sleep is polysomnography (PSG), but it can be obtrusive and costly. Actigraphy is a relatively low-cost and unobtrusive alternative to PSG. Of particular interest in measuring sleep from actigraphy is prediction of sleep-wake states. Current literature on prediction of sleep-wake states from actigraphy consists of methods that use population data, which we call generalized models. However, accounting for variability of sleep patterns across individuals calls for personalized models of sleep-wake states prediction that could be potentially better suited to individual-level data and yield more accurate estimation of sleep.

PURPOSE

To investigate the validity of developing personalized machine learning models, trained and tested on individual-level actigraphy data, for improved prediction of sleep-wake states and reliable estimation of nightly sleep parameters.

PARTICIPANTS AND METHODS

We used a dataset including 54 participants and systematically trained and tested 5 different personalized machine learning models as well as their generalized counterparts. We evaluated model performance compared to concurrent PSG through extensive machine learning experiments and statistical analyses.

RESULTS

Our experiments show the superiority of personalized models over their generalized counterparts in estimating PSG-derived sleep parameters. Personalized models of regularized logistic regression, random forest, adaptive boosting, and extreme gradient boosting achieve estimates of total sleep time, wake after sleep onset, sleep efficiency, and number of awakenings that are closer to those obtained by PSG, in absolute difference, than the same estimates from their generalized counterparts. We further show that the difference between estimates of sleep parameters obtained by personalized models and those of PSG is statistically non-significant.

CONCLUSION

Personalized machine learning models of sleep-wake states outperform their generalized counterparts in terms of estimating sleep parameters and are indistinguishable from PSG labeled sleep-wake states. Personalized machine learning models can be used in actigraphy studies of sleep health and potentially screening for some sleep disorders.

Improving Sleep Quality Assessment Using Wearable Sensors by Including Information From Postural/Sleep Position Changes and Body Acceleration: A Comparison of Chest-Worn Sensors, Wrist Actigraphy, and Polysomnography

STUDY OBJECTIVES

To improve sleep quality assessment using a single chest-worn sensor by extracting body acceleration and sleep position changes.

METHODS

Sleep patterns of 21 participants (50.8 ± 12.8 years, 47.8% female) with self-reported sleep problems were simultaneously recorded using a chest sensor (Chest), an Actiwatch (Wrist), and polysomnography (PSG) during overnight sleep laboratory assessment. An algorithm for Chest was developed to detect sleep/wake epochs based on body acceleration and sleep position/postural changes data, which were then used to estimate sleep parameters of interest. Comparisons between Chest and Wrist with respect to PSG were performed. Identification of sleep/wake epochs was assessed by estimating sensitivity, specificity, and accuracy. Agreement between sensor-derived sleep parameters and PSG was assessed using correlation coefficients and Bland-Altman analysis.

RESULTS

Chest identified sleep/wake epochs with an accuracy of on average 6% higher than Wrist (85.8% versus 79.8%). Similar trends were observed for sensitivity/specificity values. Correlation between Wrist and PSG was poor for most of the sleep parameters of interest (r = 0.0-0.3); however, Chest and PSG correlation showed moderate to strong agreement (r = 0.4-0.8) with relatively low bias and high precision bias (precision): 9.2 (13.2) minutes for sleep onset latency; 17.3(34.8) minutes for total sleep time; 7.5 (29.8) minutes for wake after sleep onset; and 2.0 (7.3)% for sleep efficacy.

CONCLUSIONS

Combination of sleep postural/position changes and body acceleration improved detection of sleep/wake epochs compared to wrist acceleration alone. The chest sensors also improved estimation of sleep parameters of interest with stronger agreement with PSG. Our findings may expand the application of wearable sensors to clinically assess sleep outside of a sleep laboratory.