A classification approach to estimating human circadian phase under circadian alignment from actigraphy and photometry data

Abstract

The time of dim light melatonin onset (DLMO) is the gold standard for circadian phase assessment in humans, but collection of samples for DLMO is time and resourceintensive. Numerous studies have attempted to estimate circadian phase from actigraphy data, but most of these studies have involved individuals on controlled and stable sleep-wake schedules, with mean errors reported between 0.5 and 1 hour. We found that such algorithms are less successful in estimating DLMO in a population of college students with more irregular schedules: Mean errors in estimating the time of DLMO are approximately 1.5-1.6 hours. We reframed the problem as a classification problem and estimated whether an individual's current phase was before or after DLMO. Using a neural network, we found high classification accuracy of about 90%, which decreased the mean error in DLMO estimation-identifying the time at which the switch in classification occurs-to approximately 1.3 hours. To test whether this classification approach was valid when activity and circadian rhythms are decoupled, we applied the same neural network to data from inpatient forced desynchrony studies in which participants are scheduled to sleep and wake at all circadian phases (rather than their habitual schedules). In participants on forced desynchrony protocols, overall classification accuracy dropped to 55%-65% with a range of 20%-80% for a given day; this accuracy was highly dependent upon the phase angle (ie, time) between DLMO and sleep onset, with the highest accuracy at phase angles associated with nighttime sleep. Circadian patterns in activity, therefore, should be included 2 of 13 | BROWN et al.

Distinct Circadian Assessments From Wearable Data Reveal Social Distancing Promoted Internal Desynchrony Between Circadian Markers

“…This model has been validated against multiple carefully controlled laboratory studies ( 35 , 36 ), and light level (in lux) is the only input in the original design of the model. However, recent advances in the field have shown that incorporating activity is necessary to predict the circadian phase in the field setting ( 5 , 12 , 37 ). In particular, activity is correlated with light, and it has been shown that replacing light by activity levels measured from wearable devices coupled with the mathematical model can predict circadian phase with an error within 1 h for people living in regular conditions ( 5 , 12 ), Initial conditions were generated from a limit-cycle [as in Huang et al ( 5 )]: the start point was determined to be 35 days prior to the social distancing date, and the model was simulated until 35 days after social distancing.…”

The Next Generation of Fatigue Prediction Models: Evaluating Current Trends in Biomathematical Modelling for Safety Optimization

“…For instance, recent circadian modelling research has indicated that lighting conditions bear strong predictive influence over circadian angle of entrainment and preferred sleep timing (Papatsimpa et al, 2021;Phillips et al, 2019). Allowing BMM circadian phase parameters to be informed from actigraphy and photometry data passively (see Brown et al, 2021) would reduce model complexity and improve predictive accuracy.…”

CARE: a novel wearable-derived feature linking circadian amplitude to human cognitive functions

“…Rhythmic fluctuations driven by the core clock could be illustrated as a sinusoidal curve with a period of about 24 hours [8], and hence the circadian rhythms are typically described using three key parameters (Supplementary Figure 1): period (the length of one cycle), phase (the timing of a reference point in the cycle relative to a fixed event), and amplitude (the difference between crest and trough values of the cycle) [9]. While methods for measuring period [10,11] and phase [12][13][14][15][16][17] are well-established and accurate in both laboratory and field studies, assessing circadian amplitude remains challenging.…”