Automatic sleeping time estimation and mild traumatic brain injury (mTBI) detection using actigraphy data. (April 2021)

• Record Type: Journal Article
• Authors: Yin, Jiawei; Julius, A. Agung; Wen, John T.; Hanifin, John P.; Warfield, Benjamin; Brainard, George C.
• Is Part Of: Biomedical signal processing and control. Volume 66(2021)
• Publication Date: 2021-04
• Subjects: Actigraphy -- Circadian rhythm -- Sleep -- Traumatic brain injury (TBI) -- Concussion detection; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2021.102430
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Actigraphy is used to estimate sleep/wake state and sleep quality features. Variations in sleep quality features and circadian rhythmicity can be used to detect concussion. The resulting Bayes classifier for concussion detection achieves up to 90% accuracy. Abstract: Sleep schedule and circadian phase irregularity are associated with some health problems and diseases, e.g., narcolepsy, circadian disorder, and concussion. Actigraphy has been widely used in the study of sleep and circadian rhythms. This paper presents a method for estimating the sleep/wake state based on the minute-by-minute actigraphy data measured by wrist actigraphy and its associated scoring software. The circadian phase shift is estimated from actigraphy data using an adaptive notch filter algorithm. Compared with the scoring methods used in wrist actigraphy, the estimated wake-onset time from our method is more consistent with the sleep logs reported by the subjects. Concussion detection based on the sleep-related features calculated from actigraphy data is also presented, with detection accuracy up to about 90%. This result implies that the concussion is closely related to sleep and circadian disruption.