Abstract
Circadian rhythms influence multiple essential biological activities, including sleep, performance, and mood. The dim light melatonin onset (DLMO) is the gold standard for measuring human circadian phase (i.e., timing). The collection of DLMO is expensive and time consuming since multiple saliva or blood samples are required overnight in special conditions, and the samples must then be assayed for melatonin. Recently, several computational approaches have been designed for estimating DLMO. These methods collect daily sampled data (e.g., sleep onset/offset times) or frequently sampled data (e.g., light exposure/skin temperature/physical activity collected every minute) to train learning models for estimating DLMO. One limitation of these studies is that they only leverage one time-scale data. We propose a two-step framework for estimating DLMO using data from both time scales. The first step summarizes data from before the current day, whereas the second step combines this summary with frequently sampled data of the current day. We evaluate three moving average models that input sleep timing data as the first step and use recurrent neural network models as the second step. The results using data from 207 undergraduates show that our two-step model with two time-scale features has statistically significantly lower root-mean-square errors than models that use either daily sampled data or frequently sampled data.
- Kelly Glazer Baron and Kathryn J. Reid. 2014. Circadian misalignment and health. International Review of Psychiatry 26, 2 (2014), 139–154.Google Scholar
- Susan Benloucif, Helen J. Burgess, Elizabeth B. Klerman, Alfred J. Lewy, Benita Middleton, Patricia J. Murphy, Barbara L. Parry, and Victoria L. Revell. 2008. Measuring melatonin in humans. Journal of Clinical Sleep Medicine 4, 01 (2008), 66–69.Google Scholar
- Maria A. Bonmati-Carrion, Benita Middleton, Victoria L. Revell, Debra J. Skene, Angeles Rol, and Juan A. Madrid. 2014. Circadian phase assessment by ambulatory monitoring in humans: Correlation with dim light melatonin onset. Chronobiology International 31, 1 (2014), 37–51.Google Scholar
- Lindsey S. Brown, Melissa A. St. Hilaire, Andrew W. McHill, Andrew J. K. Phillips, Laura K. Barger, Akane Sano, Rosalind W. Picard, Elizabeth B. Klerman, and Francis J. Doyle III. 2018. A neural network predicts human circadian phase from non-invasive, short-timeframe actigraphy and demographic data: A step towards automated control of circadian phase. Poster presented at the 2018 Society for Research on Biological Rhythms Conference.Google Scholar
- Helen J. Burgess and Charmane I. Eastman. 2005. The dim light melatonin onset following fixed and free sleep schedules. Journal of Sleep Research 14, 3 (2005), 229–237.Google Scholar
- Helen J. Burgess and Louis F. Fogg. 2008. Individual differences in the amount and timing of salivary melatonin secretion. PLoS ONE 3, 8 (2008), e3055.Google Scholar
- Tina M. Burke, Rachel R. Markwald, Andrew W. McHill, Evan D. Chinoy, Jesse A. Snider, Sara C. Bessman, Christopher M. Jung, John S. O’Neill, and Kenneth P. Wright. 2015. Effects of caffeine on the human circadian clock in vivo and in vitro. Science Translational Medicine 7, 305 (2015), 305ra146.Google Scholar
- Christian Cajochen, Sylvia Frey, Doreen Anders, Jakub Späti, Matthias Bues, Achim Pross, Ralph Mager, Anna Wirz-Justice, and Oliver Stefani. 2011. Evening exposure to a light emitting diodes (LED)-backlit computer screen affects circadian physiology and cognitive performance. American Journal of Physiology-Heart and Circulatory Physiology 110, 5 (2011), 1432–1438.Google Scholar
- Anne-Marie Chang, Daniel Aeschbach, Jeanne F. Duffy, and Charles A. Czeisler. 2015. Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness. Proceedings of the National Academy of Sciences 112, 4 (2015), 1232–1237.Google Scholar
Cross Ref
- Junyoung Chung, Caglar Gulcehre, KyungHyun Cho, and Yoshua Bengio. 2014. Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv:1412.3555Google Scholar
- Stephanie J. Crowley, Christine Acebo, Gahan Fallone, and Mary A. Carskadon. 2006. Estimating dim light melatonin onset (DLMO) phase in adolescents using summer or school-year sleep/wake schedules. Sleep 29, 12 (2006), 1632–1641.Google Scholar
- Charles A. Czeisler and Orfeu M. Buxton. 2017. The human circadian timing system and sleep-wake regulation. In Principles and Practice of Sleep Medicine (6th ed.). Elsevier, 362–376.Google Scholar
- Charles A. Czeisler, Michael P. Johnson, Jeanne F. Duffy, Emery N. Brown, Joseph M. Ronda, and Richard E. Kronauer. 1990. Exposure to bright light and darkness to treat physiologic maladaptation to night work. New England Journal of Medicine 322, 18 (1990), 1253–1259.Google Scholar
Cross Ref
- Simon L. Davies, Andrew A. Neath, and Joseph E. Cavanaugh. 2005. Cross validation model selection criteria for linear regression based on the Kullback–Leibler discrepancy. Statistical Methodology 2, 4 (2005), 249–266.Google Scholar
Cross Ref
- Luigi Di Lorenzo, Giovanni De Pergola, Carlo Zocchetti, Nicola L’Abbate, Antonella Basso, Nicola Pannacciulli, Mauro Cignarelli, Riccardo Giorgino, and Leonardo Soleo. 2003. Effect of shift work on body mass index: Results of a study performed in 319 glucose-tolerant men working in a southern Italian industry. International Journal of Obesity 27, 11 (2003), 1353.Google Scholar
- Tianxiang Gao and Vladimir Jojic. 2016. Degrees of freedom in deep neural networks. arXiv:1603.09260Google Scholar
- Enrique A. Gil, Xavier L. Aubert, Els I. S. Møst, and Domien G. M. Beersma. 2013. Human circadian phase estimation from signals collected in ambulatory conditions using an autoregressive model. Journal of Biological Rhythms 28, 2 (2013), 152–163.Google Scholar
- Samuel B. Green. 1991. How many subjects does it take to do a regression analysis? Multivariate Behavioral Research 26, 3 (1991), 499–510.Google Scholar
Cross Ref
- Simon Haykin. 1994. Neural Networks. Vol. 2. Prentice Hall, New York, NY.Google Scholar
- Melissa A. St. Hilaire, Claude Gronfier, Jamie M. Zeitzer, and Elizabeth B. Klerman. 2007. A physiologically based mathematical model of melatonin including ocular light suppression and interactions with the circadian pacemaker. Journal of Pineal Research 43, 3 (2007), 294–304.Google Scholar
- Melissa A. St. Hilaire, Elizabeth B. Klerman, Sat Bir S. Khalsa, Kenneth P. Wright Jr., Charles A. Czeisler, and Richard E. Kronauer. 2007. Addition of a non-photic component to a light-based mathematical model of the human circadian pacemaker. Journal of Theoretical Biology 247, 4 (2007), 583–599.Google Scholar
- Sat Bir S. Khalsa, Megan E. Jewett, Christian Cajochen, and Charles A. Czeisler. 2003. A phase response curve to single bright light pulses in human subjects. Journal of Physiology 549, 3 (2003), 945–952.Google Scholar
- Elizabeth B. Klerman, Hayley B. Gershengorn, Jeanne F. Duffy, and Richard E. Kronauer. 2002. Comparisons of the variability of three markers of the human circadian pacemaker. Journal of Biological Rhythms 17, 2 (2002), 181–193.Google Scholar
- Vitaliy Kolodyazhniy, Jakub Späti, Sylvia Frey, Thomas Götz, Anna Wirz-Justice, Kurt Kräuchi, Christian Cajochen, and Frank H. Wilhelm. 2011. Estimation of human circadian phase via a multi-channel ambulatory monitoring system and a multiple regression model. Journal of Biological Rhythms 26, 1 (2011), 55–67.Google Scholar
Cross Ref
- Vitaliy Kolodyazhniy, Jakub Späti, Sylvia Frey, Thomas Götz, Anna Wirz-Justice, Kurt Kräuchi, Christian Cajochen, and Frank H. Wilhelm. 2012. An improved method for estimating human circadian phase derived from multichannel ambulatory monitoring and artificial neural networks. Chronobiology International 29, 8 (2012), 1078–1097.Google Scholar
Cross Ref
- Richard E. Kronauer, Daniel B. Forger, and Megan E. Jewett. 1999. Quantifying human circadian pacemaker response to brief, extended, and repeated light stimuli over the phototopic range. Journal of Biological Rhythms 14, 6 (1999), 501–516.Google Scholar
Cross Ref
- Stacia K. Martin and Charmane I. Eastman. 2002. Sleep logs of young adults with self-selected sleep times predict the dim light melatonin onset. Chronobiology International 19, 4 (2002), 695–707.Google Scholar
- Andrew W. McHill, Andrew J. K. Phillips, Charles A. Czeisler, Leigh Keating, Karen Yee, Laura K. Barger, Marta Garaulet, Frank A. J. L. Scheer, and Elizabeth B. Klerman. 2017. Later circadian timing of food intake is associated with increased body fat. American Journal of Clinical Nutrition 106, 5 (2017), 1213–1219.Google Scholar
- Tomáš Mikolov, Martin Karafiát, Lukáš Burget, Jan Černockỳ, and Sanjeev Khudanpur. 2010. Recurrent neural network based language model. In Proceedings of the 11th Annual Conference of the International Speech Communication Association.Google Scholar
Cross Ref
- Ralph E. Mistlberger. 2005. Circadian regulation of sleep in mammals: Role of the suprachiasmatic nucleus. Brain Research Reviews 49, 3 (2005), 429–454.Google Scholar
Cross Ref
- Christopher J. Morris, Taylor E. Purvis, Kun Hu, and Frank A. J. L. Scheer. 2016. Circadian misalignment increases cardiovascular disease risk factors in humans. Proceedings of the National Academy of Sciences 113, 10 (2016), e1402–e1411.Google Scholar
- Noboru Murata, Shuji Yoshizawa, and Shun-Ichi Amari. 1994. Network information criterion-determining the number of hidden units for an artificial neural network model. IEEE Transactions on Neural Networks 5, 6 (1994), 865–872.Google Scholar
Digital Library
- Seithikurippu R. Pandi-Perumal, Marcel Smits, Warren Spence, Venkataramanujan Srinivasan, Daniel P. Cardinali, Alan D. Lowe, and Leonid Kayumov. 2007. Dim light melatonin onset (DLMO): A tool for the analysis of circadian phase in human sleep and chronobiological disorders. Progress in Neuro-Psychopharmacology and Biological Psychiatry 31, 1 (2007), 1–11.Google Scholar
Cross Ref
- Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, et al. 2019. PyTorch: An imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems 32, H. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. Fox, and R. Garnett (Eds.). Curran Associates., 8024–8035. http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf.Google Scholar
- Andrew J. K. Phillips, William M. Clerx, Conor S. O’Brien, Akane Sano, Laura K. Barger, Rosalind W. Picard, Steven W. Lockley, Elizabeth B. Klerman, and Charles A. Czeisler. 2017. Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing. Scientific Reports 7, 1 (2017), 3216.Google Scholar
Cross Ref
- Payam Refaeilzadeh, Lei Tang, and Huan Liu. 2007. On comparison of feature selection algorithms. In Proceedings of the AAAI Workshop on Evaluation Methods for Machine Learning II, Vol. 3. 5.Google Scholar
- Benjamin Rusak and Irving Zucker. 1979. Neural regulation of circadian rhythms. Physiological Reviews 59, 3 (1979), 449–526.Google Scholar
- Robert L. Sack, Dennis Auckley, R. Robert Auger, Mary A. Carskadon, Kenneth P. Wright Jr., Michael V. Vitiello, and Irina V. Zhdanova. 2007. Circadian rhythm sleep disorders: Part I, basic principles, shift work and jet lag disorders. Sleep 30, 11 (2007), 1460–1483.Google Scholar
Cross Ref
- Akane Sano. 2016. Measuring College Students’ Sleep, Stress, Mental Health and Wellbeing with Wearable Sensors and Mobile Phones. Ph.D. Dissertation. Massachusetts Institute of Technology.Google Scholar
- Akane Sano, Weixuan Chen, Daniel Lopez-Martinez, Sara Taylor, and Rosalind W. Picard. 2019. Multimodal ambulatory sleep detection using LSTM recurrent neural networks. IEEE Journal of Biomedical and Health Informatics 23, 4 (2019), 1607–1617.Google Scholar
Cross Ref
- Akane Sano, Sara Taylor, Andrew W. McHill, Andrew J. K. Phillips, Laura K. Barger, Elizabeth B. Klerman, and Rosalind Picard. 2018. Identifying objective physiological markers and modifiable behaviors for self-reported stress and mental health status using wearable sensors and mobile phones: Observational study. Journal of Medical Internet Research 20, 6 (2018), e210.Google Scholar
Cross Ref
- Pawel Smialowski, Dmitrij Frishman, and Stefan Kramer. 2009. Pitfalls of supervised feature selection. Bioinformatics 26, 3 (2009), 440–443.Google Scholar
- Julia E. Stone, Andrew J. K. Phillips, Suzanne Ftouni, Michelle Magee, Mark Howard, Steven W. Lockley, Tracey L. Sletten, Clare Anderson, Shantha M. W. Rajaratnam, and Svetlana Postnova. 2019. Generalizability of a neural network model for circadian phase prediction in real-world conditions. Scientific Reports 9, 1 (2019), 1–17.Google Scholar
- Sara Ann Taylor, Natasha Jaques, Ehimwenma Nosakhare, Akane Sano, and Rosalind W. Picard. 2020. Personalized multitask learning for predicting tomorrow’s mood, stress, and health. IEEE Transactions on Affective Computing 11, 2 (2020), 200–213.Google Scholar
Cross Ref
Index Terms
Sensor-Based Estimation of Dim Light Melatonin Onset Using Features of Two Time Scales
Recommendations
Multifractality in heartbeat dynamics in patients undergoing beating-heart myocardial revascularization
BackgroundThe multifractal approach of HRV analysis offers new insight into the mechanisms of autonomic modulation of the diseased hearts and has a potential to depict subtle changes in cardiac autonomic nervous control not revealed by conventional ...
Light-based circadian rhythm control
Light is a strong synchronizer for circadian rhythm - the 24-h biological oscillation in plants, insects, and mammals. This paper considers the circadian entrainment problem for a popular circadian oscillation model (the Kronauer model) by using light ...
Towards qualitative and quantitative prediction and detection of parturition onset in sows using light barriers
A single light barrier can be used for the prediction and detection of parturition.A validated classification performance close to 90% was achieved.Approaches for measuring the parturition related activity increase were compared.A method for a ...






Comments