High-throughput visual assessment of sleep stages in mice using machine learning

SLEEP ◽  
2021 ◽  
Author(s):  
Brian Geuther ◽  
Mandy Chen ◽  
Raymond J Galante ◽  
Owen Han ◽  
Jie Lian ◽  
...  
Keyword(s):  
Machine Learning ◽  
Eye Movement ◽  
High Throughput ◽  
Data Augmentation ◽  
Sleep Stage ◽  
Visual Assessment ◽  
Video Data ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep

Abstract Study Objectives Sleep is an important biological process that is perturbed in numerous diseases, and assessment its substages currently requires implantation of electrodes to carry out electroencephalogram/electromyogram (EEG/EMG) analysis. Although accurate, this method comes at a high cost of invasive surgery and experts trained to score EEG/EMG data. Here, we leverage modern computer vision methods to directly classify sleep substages from video data. This bypasses the need for surgery and expert scoring, provides a path to high-throughput studies of sleep in mice. Methods We collected synchronized high-resolution video and EEG/EMG data in 16 male C57BL/6J mice. We extracted features from the video that are time and frequency-based and used the human expert-scored EEG/EMG data to train a visual classifier. We investigated several classifiers and data augmentation methods. Results Our visual sleep classifier proved to be highly accurate in classifying wake, non-rapid eye movement sleep (NREM), and rapid eye movement sleep (REM) states, and achieves an overall accuracy of 0.92 +/- 0.05 (mean +/- SD). We discover and genetically validate video features that correlate with breathing rates, and show low and high variability in NREM and REM sleep, respectively. Finally, we apply our methods to non-invasively detect that sleep stage disturbances induced by amphetamine administration. Conclusions We conclude that machine learning based visual classification of sleep is a viable alternative to EEG/EMG based scoring. Our results will enable non-invasive high-throughput sleep studies and will greatly reduce the barrier to screening mutant mice for abnormalities in sleep.

Influence of sleep on tensor palatini EMG and upper airway resistance in normal men

1991 ◽  
Vol 70 (6) ◽  
pp. 2574-2581 ◽  
Author(s):  
D. J. Tangel ◽  
W. S. Mezzanotte ◽  
D. P. White
Keyword(s):  
Eye Movement ◽  
Airway Resistance ◽  
Sleep Stage ◽  
Upper Airway ◽  
Sleep Stages ◽  
Normal Male ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep ◽  
Airway Narrowing ◽  
Upper Airway Resistance

We propose that a sleep-induced decrement in the activity of the tensor palatini (TP) muscle could induce airway narrowing in the area posterior to the soft palate and therefore lead to an increase in upper airway resistance in normal subjects. We investigated the TP to determine the influence of sleep on TP muscle activity and the relationship between changing TP activity and upper airway resistance over the entire night and during short sleep-awake transitions. Seven normal male subjects were studied on a single night with wire electrodes placed in both TP muscles. Sleep stage, inspiratory airflow, transpalatal pressure, and TP moving time average electromyogram (EMG) were continuously recorded. In addition, in two of the seven subjects the activity (EMG) of both the TP and the genioglossus muscle simultaneously was recorded throughout the night. Upper airway resistance increased progressively from wakefulness through the various non-rapid-eye-movement sleep stages, as has been previously described. The TP EMG did not commonly demonstrate phasic activity during wakefulness or sleep. However, the tonic EMG decreased progressively and significantly (P less than 0.05) from wakefulness through the non-rapid-eye-movement sleep stages [awake, 4.6 +/- 0.3 (SE) arbitrary units; stage 1, 2.6 +/- 0.3; stage 2, 1.7 +/- 0.5; stage 3/4, 1.5 +/- 0.8]. The mean correlation coefficient between TP EMG and upper airway resistance across all sleep states was (-0.46). This mean correlation improved over discrete sleep-awake transitions (-0.76). No sleep-induced decrement in the genioglossus activity was observed in the two subjects studied.(ABSTRACT TRUNCATED AT 250 WORDS)

Metabolic rate and breathing during sleep

1985 ◽  
Vol 59 (2) ◽  
pp. 384-391 ◽  
Author(s):  
D. P. White ◽  
J. V. Weil ◽  
C. W. Zwillich
Keyword(s):  
Metabolic Rate ◽  
Eye Movement ◽  
Sleep Stage ◽  
Minute Ventilation ◽  
Respiratory Control ◽  
Normal Subjects ◽  
Co2 Production ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep

Recent investigation suggests that both ventilation (VE) and the chemical sensitivity of the respiratory control system correlate closely with measures of metabolic rate [O2 consumption (VO2) and CO2 production (VCO2)]. However, these associations have not been carefully investigated during sleep, and what little information is available suggests a deterioration of the relationships. As a result we measured VE, ventilatory pattern, VO2, and VCO2 during sleep in 21 normal subjects (11 males and 10 females) between the ages of 21 and 77 yr. When compared with values for awake subjects, expired ventilation decreased 8.2 +/- 2.3% (SE) during sleep and was associated with a 8.5 +/- 1.6% decrement in VO2 and a 12.3 +/- 1.7% reduction in VCO2, all P less than 0.01. The decrease in ventilation was a product primarily of a significant decrease in tidal volume with little change in frequency. None of these findings were dependent on sleep stage with results in rapid-eye-movement (REM) and non-rapid-eye-movement sleep being similar. Through all sleep stages ventilation remained tightly correlated with VO2 and VCO2 both within a given individual and between subjects. Although respiratory rhythmicity was somewhat variable during REM sleep, minute ventilation continued to correlate with VO2 and VCO2. None of the parameters described above were influenced by age or gender, with male and female subjects demonstrating similar findings. Ten of the subjects demonstrated at least occasional apneas. These individuals, however, were not found to differ from those without apnea in any other measure of ventilation or metabolic rate.

Partitioning of inhaled ventilation between the nasal and oral routes during sleep in normal subjects

2003 ◽  
Vol 94 (3) ◽  
pp. 883-890 ◽  
Author(s):  
Michael F. Fitzpatrick ◽  
Helen S. Driver ◽  
Neela Chatha ◽  
Nha Voduc ◽  
Alison M. Girard
Keyword(s):  
Eye Movement ◽  
Slow Wave ◽  
Sleep Stage ◽  
Minute Ventilation ◽  
Normal Subjects ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Nasal Mask ◽  
Rapid Eye Movement Sleep ◽  
Significant Difference

The oral and nasal contributions to inhaled ventilation were simultaneously quantified during sleep in 10 healthy subjects (5 men, 5 women) aged 43 ± 5 yr, with normal nasal resistance (mean 2.0 ± 0.3 cmH2O · l−1 · s−1) by use of a divided oral and nasal mask. Minute ventilation awake (5.9 ± 0.3 l/min) was higher than that during sleep (5.2 ± 0.3 l/min; P < 0.0001), but there was no significant difference in minute ventilation between different sleep stages ( P = 0.44): stage 2 5.3 ± 0.3, slow-wave 5.2 ± 0.2, and rapid-eye-movement sleep 5.2 ± 0.2 l/min. The oral fraction of inhaled ventilation during wakefulness (7.6 ± 4%) was not significantly different from that during sleep (4.3 ± 2%; mean difference 3.3%, 95% confidence interval −2.1–8.8%, P = 0.19), and no significant difference ( P = 0.14) in oral fraction was observed between different sleep stages: stage two 5.1 ± 2.8, slow-wave 4.2 ± 1.8, rapid-eye-movement 3.1 ± 1.7%. Thus the inhaled oral fraction in normal subjects is small and does not change significantly with sleep stage.

Heat production during sleep

1984 ◽  
Vol 56 (3) ◽  
pp. 671-677 ◽  
Author(s):  
C. M. Shapiro ◽  
C. C. Goll ◽  
G. R. Cohen ◽  
I. Oswald
Keyword(s):  
Eye Movement ◽  
Heat Production ◽  
Indirect Calorimetry ◽  
Sleep Stage ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep ◽  
Average Heat ◽  
Stage 4

Heat production during sleep was studied by continuous indirect calorimetry with simultaneous electroencephalographic monitoring. Controlling for gross influences on heat production, comparisons of heat production during different sleep stages showed heat production in stage 4 sleep to be significantly lower than in other sleep stages. There appeared to be a gradation in heat production in non-rapid-eye-movement stages of sleep with stage 2 higher and stage 4 lower than stage 3. Heat production in stage 4 was less variable than in any other sleep stage. Both the level and variability of heat production was similar in stage 2 and rapid-eye-movement sleep. Heat production during the night was 9% lower than during resting wakefulness. The average heat production in stage 4 sleep was 14.4% lower than resting wakeful values.

scholarly journals Dream Recall upon Awakening from Non-Rapid Eye Movement Sleep in Older Adults: Electrophysiological Pattern and Qualitative Features

2020 ◽  
Vol 10 (6) ◽  
pp. 343 ◽  
Author(s):  
Serena Scarpelli ◽  
Aurora D’Atri ◽  
Chiara Bartolacci ◽  
Maurizio Gorgoni ◽  
Anastasia Mangiaruga ◽  
...  
Keyword(s):  
Eye Movement ◽  
Sleep Stage ◽  
Nrem Sleep ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep ◽  
Dream Recall ◽  
Cortical Arousal ◽  
Beta Power ◽  
Dream Report

Several findings support the activation hypothesis, positing that cortical arousal promotes dream recall (DR). However, most studies have been carried out on young participants, while the electrophysiological (EEG) correlates of DR in older people are still mostly unknown. We aimed to test the activation hypothesis on 20 elders, focusing on the Non-Rapid Eye Movement (NREM) sleep stage. All the subjects underwent polysomnography, and a dream report was collected upon their awakening from NREM sleep. Nine subjects were recallers (RECs) and 11 were non-RECs (NRECs). The delta and beta EEG activity of the last 5 min and the total NREM sleep was calculated by Fast Fourier Transform. Statistical comparisons (RECs vs. NRECs) revealed no differences in the last 5 min of sleep. Significant differences were found in the total NREM sleep: the RECs showed lower delta power over the parietal areas than the NRECs. Consistently, statistical comparisons on the activation index (delta/beta power) revealed that RECs showed a higher level of arousal in the fronto-temporal and parieto-occipital regions than NRECs. Both visual vividness and dream length are positively related to the level of activation. Overall, our results are consistent with the view that dreaming and the storage of oneiric contents depend on the level of arousal during sleep, highlighting a crucial role of the temporo-parietal-occipital zone.

scholarly journals Why Are Seizures Rare in Rapid Eye Movement Sleep? Review of the Frequency of Seizures in Different Sleep Stages

2013 ◽  
Vol 2013 ◽  
pp. 1-10 ◽  
Author(s):  
Marcus Ng ◽  
Milena Pavlova
Keyword(s):  
Eye Movement ◽  
Rem Sleep ◽  
Low Voltage ◽  
Sleep Stage ◽  
Sleep Stages ◽  
Seizure Freedom ◽  
Rapid Eye Movement ◽  
Focal Seizures ◽  
Generalized Seizures ◽  
Interictal Discharges

Since the formal characterization of sleep stages, there have been reports that seizures may preferentially occur in certain phases of sleep. Through ascending cholinergic connections from the brainstem, rapid eye movement (REM) sleep is physiologically characterized by low voltage fast activity on the electroencephalogram, REMs, and muscle atonia. Multiple independent studies confirm that, in REM sleep, there is a strikingly low proportion of seizures (~1% or less). We review a total of 42 distinct conventional and intracranial studies in the literature which comprised a net of 1458 patients. Indexed to duration, we found that REM sleep was the most protective stage of sleep against focal seizures, generalized seizures, focal interictal discharges, and two particular epilepsy syndromes. REM sleep had an additional protective effect compared to wakefulness with an average 7.83 times fewer focal seizures, 3.25 times fewer generalized seizures, and 1.11 times fewer focal interictal discharges. In further studies REM sleep has also demonstrated utility in localizing epileptogenic foci with potential translation into postsurgical seizure freedom. Based on emerging connectivity data in sleep, we hypothesize that the influence of REM sleep on seizures is due to a desynchronized EEG pattern which reflects important connectivity differences unique to this sleep stage.

First rapid eye movement sleep periods and sleep-onset rapid eye movement periods in sleep-stage sequencing of hypersomnias

2013 ◽  
Vol 14 (9) ◽  
pp. 897-901 ◽  
Author(s):  
Panagis Drakatos ◽  
Christopher A. Kosky ◽  
Sean E. Higgins ◽  
Rexford T. Muza ◽  
Adrian J. Williams ◽  
...  
Keyword(s):  
Eye Movement ◽  
Sleep Stage ◽  
Sleep Onset ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep

scholarly journals Regulation of REM Sleep by Inhibitory Neurons in the Dorsomedial Medulla

2020 ◽  
Author(s):  
Joseph A. Stucynski ◽  
Amanda L. Schott ◽  
Justin Baik ◽  
Shinjae Chung ◽  
Franz Weber
Keyword(s):  
Eye Movement ◽  
Rem Sleep ◽  
Gabaergic Neurons ◽  
Inhibitory Neurons ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep ◽  
Brain Rhythms ◽  
Slow Activity ◽  
Raphé Nuclei

ABSTRACTThe two major stages of mammalian sleep – rapid eye movement sleep (REMs) and non-REM sleep (NREMs) – are characterized by distinct brain rhythms ranging from millisecond to minute-long (infraslow) oscillations. The mechanisms controlling transitions between sleep stages and how they are synchronized with infraslow rhythms remain poorly understood. Using opto- and chemogenetic manipulation, we show that GABAergic neurons in the dorsomedial medulla (dmM) promote the initiation and maintenance of REMs, in part through their projections to the dorsal and median raphe nuclei. Fiber photometry revealed that dmM GABAergic neurons are strongly activated during REMs. During NREMs, their activity fluctuated in close synchrony with infraslow oscillations in the spindle band of the electroencephalogram, and the phase of this rhythm modulated the latency of optogenetically induced REMs episodes. Thus, dmM inhibitory neurons powerfully promote REMs, and their slow activity fluctuations may coordinate transitions from NREMs to REMs with infraslow brain rhythms.

scholarly journals REM sleep: unique associations with behavior, corticosterone regulation and apoptotic pathways in chronic stress in mice

2018 ◽  
Author(s):  
Mathieu Nollet ◽  
Harriet Hicks ◽  
Andrew P. McCarthy ◽  
Huihai Wu ◽  
Carla S. Möller-Levet ◽  
...  
Keyword(s):  
Machine Learning ◽  
Eye Movement ◽  
Chronic Stress ◽  
Rem Sleep ◽  
Chronic Mild Stress ◽  
Specific Aspect ◽  
Mild Stress ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep ◽  
Apoptotic Pathways

AbstractOne of sleep’s putative functions is mediation of adaptation to waking experiences. Chronic stress is a common waking experience, however, which specific aspect of sleep is most responsive, and how sleep changes relate to behavioral disturbances and molecular correlates remain unknown. We quantified sleep, physical, endocrine and behavioral variables and the brain and blood transcriptome in mice exposed to nine weeks of unpredictable chronic mild stress (UCMS). Comparing 46 phenotypical variables revealed that rapid-eye-movement sleep (REMS), corticosterone regulation and coat state were most responsive to UCMS. REMS theta oscillations were enhanced whereas delta oscillations in non-REMS were unaffected. Transcripts affected by UCMS in the prefrontal cortex, hippocampus, hypothalamus and blood were associated with inflammatory and immune responses. A machine learning approach controlling for unspecific UCMS effects identified transcriptomic predictors for specific phenotypes and their overlap. Transcriptomic predictor sets for the inter-individual variation in REMS continuity and theta activity shared many pathways with corticosterone regulation and in particular pathways implicated in apoptosis, including mitochondrial pathways. Predictor sets for REMS and anhedonia, one of the behavioral changes following UCMS, shared pathways involved in oxidative stress, cell proliferation and apoptosis. RNA predictor sets for non-NREMS parameters showed no overlap with other phenotypes. These novel data identify REMS as a core and early element of the response to chronic stress, and identify apoptotic pathways as a putative mechanism by which REMS mediates adaptation to stressful waking experiences.Significance StatementSleep is responsive to experiences during wakefulness and is altered in stress-related disorders. Whether sleep changes primarily concern rapid-eye-movement sleep (REMS) or non-REM sleep, and how they correlate with stress hormones, behavioral and transcriptomic responses remained unknown. We demonstrate using unpredictable chronic (9-weeks) mild stress that REMS is the most responsive of all the measured sleep characteristics, and correlates with deficiency in corticosterone regulation. An unbiased machine learning, controlling for unspecific effects of stress, revealed that REMS correlated with RNA predictor sets enriched in apoptosis including mitochondrial pathways. Several pathways were shared with predictors of corticosterone and behavioral responses. This unbiased approach point to apoptosis as a molecular mechanism by which REMS mediates adaptation to an ecologically relevant waking experience.

scholarly journals Remifentanil Inhibits Rapid Eye Movement Sleep but Not the Nocturnal Melatonin Surge in Humans

2008 ◽  
Vol 108 (4) ◽  
pp. 627-633 ◽  
Author(s):  
Christopher P. Bonafide ◽  
Natalie Aucutt-Walter ◽  
Nicole Divittore ◽  
Tonya King ◽  
Edward O. Bixler ◽  
...  
Keyword(s):  
Plasma Concentration ◽  
Sleep Disturbance ◽  
Eye Movement ◽  
Repeated Measures ◽  
Slow Wave Sleep ◽  
Constant Infusion ◽  
Sleep Stages ◽  
Rapid Eye Movement ◽  
Rapid Eye Movement Sleep ◽  
Placebo Capsule

Background Postoperative patients are sleep deprived. Opioids, commonly administered for postoperative pain control, are often mistakenly considered inducers of naturally occurring sleep. This study describes the effect of the opioid remifentanil on nocturnal sleep in healthy volunteers. In addition, this study tests the hypothesis that opioid-induced sleep disturbance is caused by a circadian pacemaker disturbance, reflected by suppressed nocturnal plasma concentration of melatonin. Methods Polysomnography was performed in 10 volunteers from 11:00 pm to 7:00 am for four nights at 6-day intervals. On two nights, remifentanil (0.01-0.04 microg x kg x min) was infused from 10:30 pm to 7:00 am, and either a placebo capsule or 3.0 mg melatonin was administered at 10:30 pm. On two additional nights, saline was infused, and the placebo or melatonin capsules were administered at 10:30 pm. Blood was drawn at 12:00 am, 3:00 am, and 6:00 am to measure the plasma concentration of melatonin and cortisol. A repeated-measures analysis of variance model was used to determine the effect of remifentanil on sleep stages, the effect of remifentanil on the plasma concentration of melatonin, and the effect of exogenous melatonin on remifentanil-induced sleep disturbance. Results Remifentanil inhibited rapid eye movement sleep (14.1 +/- 7.2% to 3.9 +/- 6.9%). The amount of slow wave sleep decreased from 6.8 +/- 7.6% to 3.2 +/- 6.1%, but this decrease was not statistically significant. Remifentanil did not decrease melatonin concentration. Melatonin administration did not prevent remifentanil-induced sleep disturbance. Conclusions An overnight constant infusion of remifentanil inhibits rapid eye movement sleep without suppressing the nocturnal melatonin surge.

Sign in / Sign up

Export Citation Format

Share Document