Identifying time‐resolved features of nocturnal sleep characteristics of narcolepsy using machine learning

Journal of Sleep Research, Journal Year: 2024, Volume and Issue: 33(6)

Published: April 26, 2024

Summary The differential diagnosis of narcolepsy type 1, a rare, chronic, central disorder hypersomnolence, is challenging due to overlapping symptoms with other hypersomnolence disorders. While recent years have seen significant growth in our understanding nocturnal polysomnography 1 features, there remains need for improving methods differentiate nighttime sleep features from those individuals without 1. We aimed develop machine learning framework identifying discriminate clinical controls, 2 and idiopathic hypersomnia. population included data 350 drug‐free (114 90 2, 105 hypersomnia, 41 controls) collected at the National Reference Centers Narcolepsy Montpelier, France. Several sets were explored, as well value time‐resolving architecture by analysing per quarter‐night. patterns evolution emerged that differed between increased instability observed patients Using models, we identified rapid eye movement onset best single feature distinguish By combining multiple capturing different aspects across quarter‐night periods, able further improve between‐group discrimination could identify most discriminative features. Our results highlight salient relevance assessing their time‐dependent changes during aid measure impact novel therapeutics future trials.

Language: Английский

---

Certainty about uncertainty in sleep staging: a theoretical framework

SLEEP, Journal Year: 2022, Volume and Issue: 45(8)

Published: June 8, 2022

Abstract Sleep stage classification is an important tool for the diagnosis of sleep disorders. Because staging has such a high impact on clinical outcome, it that done reliably. However, known uncertainty exists in both expert scorers and automated models. On average, agreement between human only 82.6%. In this study, we provide theoretical framework to facilitate discussion further analyses staging. To end, introduce two variants uncertainty, from statistics machine learning community: aleatoric epistemic uncertainty. We discuss what these types uncertainties are, why distinction useful, where they arise staging, recommendations how can improve future.

Language: Английский

---

U-PASS: An uncertainty-guided deep learning pipeline for automated sleep staging

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 171, P. 108205 - 108205

Published: Feb. 23, 2024

With the increasing prevalence of machine learning in critical fields like healthcare, ensuring safety and reliability these systems is crucial. Estimating uncertainty plays a vital role enhancing by identifying areas high low confidence reducing risk errors. This study introduces U-PASS, specialized human-centered pipeline tailored for clinical applications, which effectively communicates to experts collaborates with them improve predictions. U-PASS incorporates estimation at every stage process, including data acquisition, training, model deployment. Training divided into supervised pre-training step semi-supervised recording-wise finetuning step. We apply challenging task sleep staging demonstrate that it systematically improves performance stage. By optimizing training dataset, actively seeking feedback from domain informative samples, deferring most uncertain samples experts, achieves an impressive expert-level accuracy 85% on dataset elderly apnea patients. represents significant improvement over starting point 75% accuracy. The largest gain due deferral epochs expert. presents promising AI approach incorporating pipelines, improving their unlocking potential settings.

Language: Английский

---

Using Low-Cost Technology Devices for Monitoring Sleep and Environmental Factors Affecting It: A Systematic Review of the Literature

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(3), P. 1188 - 1188

Published: Jan. 24, 2025

Low-cost technology devices, such as smartphones (SPs) and smart watches (SWs), are widely used today to monitor various health effects environmental risk factors associated with them. However, the efficacy of using these devices monitoring tools is largely unknown. The present study attempts narrow this knowledge gap by reviewing recent studies in which low-cost technological were sleep factors. focuses on peer-refereed articles that appear three major scientific databases, Web Science, Scopus, ScienceDirect, published between 2002 2022. Of 15,000+ records retrieved from databases systematic literature review (PRISMA) search, 15 identified most relevant consequently analyzed. analysis shows nighttime light pollution noise commonly monitored (eight studies), followed temperature (seven humidity CO2 (four studies). In eight studies, tandems SPs SWs sleep, while six data obtained compared conventional devices. general, SP SW measurements found be fairly accurate for less noise. At same time, no conducted date analyzed demonstrated effectiveness ambient temperature, humidity, air pressure. Our general conclusion although often lack precision professional instruments, they can nevertheless large-scale field research citizen science initiatives, their feasibility several attributes have yet determined.

Language: Английский

---

Understanding Sleep Dynamics Gathered from Wearable Devices with Explainable Recurrent Neural Networks

Lecture notes in computer science, Journal Year: 2025, Volume and Issue: unknown, P. 231 - 245

Published: Jan. 1, 2025

Language: Английский

---

Sleeping With the Enemy: Drug-Resistant Epilepsy and Sleep

Epiliepsy currents/Epilepsy currents, Journal Year: 2025, Volume and Issue: unknown

Published: March 14, 2025

Factors Associated With Poor Sleep-In Children Drug-Resistant Epilepsy Proost R, Cleeren E, Jansen B, Lagae L, Van Paesschen W, K. Epilepsia . 2024;65(11):3335-3349. doi:10.1111/epi.18112. Objective: We aimed to investigate sleep in children with drug-resistant epilepsy (DRE), including developmental and epileptic encephalopathies (DEEs). Next, we examined differences macrostructure microstructure questionnaire outcomes between well-controlled (WCE) DRE. Furthermore, wanted identify factors associated poor outcome these children, as some might be targets improve neurodevelopmental outcomes. Methods: A cross-sectional study was conducted 4 18-years-old. without epilepsy, WCE, DRE were included. Overnight electroencephalography (EEG), chin electromyography electrooculography, allow staging, performed. Parents asked fill out a questionnaire. Classical five-stage scoring performed manually, spindles automatically counted, slow wave activity (SWA) the first last hour of calculated. Results: One hundred eighty-two patients included: 48 75 59 found that have significantly lower efficiency (SE%), less time spent rapid eye movement (REM) sleep, fewer spindles, SWA decline over night compared WCE. Subjectively more severe problems reported by caregivers daytime sleepiness present Least absolute shrinkage selection operator (LASSO) regression showed multifocal interictal epileptiform discharges (IEDs), benzodiazepine treatment, longer duration SE% REM time. The presence cerebral palsy spindles. Benzodiazepine drug resistance, seizures during intellectual disability, older age decline. Significance: Both are severely impacted DRE, those DEEs. parameters play distinct role disruption spindle count,

Language: Английский

---

Novel digital markers of sleep dynamics: causal inference approach revealing age and gender phenotypes in obstructive sleep apnea

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 8, 2025

Despite evidence that sleep-disorders alter sleep-stage dynamics, only a limited amount of these parameters are included and interpreted in clinical practice, mainly due to unintuitive methodologies or lacking normative values. Leveraging the matrix transition proportions, we propose (i) general framework quantify sleep-dynamics, (ii) several novel markers their alterations, (iii) demonstrate our approach using obstructive sleep apnea (OSA), one most prevalent sleep-disorder significant risk factor. Using causal inference techniques, address confounding an observational database estimate personalized by age, gender, OSA-severity. Importantly, adjusts for five categories sleep-wake-related comorbidities, factor overlooked existing research but present 48.6% OSA-subjects high-quality dataset. Key markers, such as NREM-REM-oscillations sleep-stage-specific fragmentations, were increased across all OSA-severities demographic groups. Additionally, identified distinct gender-phenotypes, suggesting females may be more vulnerable awakenings REM-sleep-disruptions. External validation on SHHS confirmed robustness detecting sleep-disordered-breathing (average AUROC = 66.4%). With advancements automated sleep-scoring wearable devices, holds promise developing low-cost screening tools sleep-, neurodegenerative-, psychiatric-disorders exhibiting altered patterns.

Language: Английский

---

Functional changes in sleep-related arousal after ketamine administration in individuals with treatment-resistant depression

Translational Psychiatry, Journal Year: 2024, Volume and Issue: 14(1)

Published: June 4, 2024

Abstract The glutamatergic modulator ketamine is associated with changes in sleep, depression, and suicidal ideation (SI). This study sought to evaluate differences arousal-related sleep metrics between 36 individuals treatment-resistant major depression (TRD) 25 healthy volunteers (HVs). It also determine whether normalizes arousal TRD ketamine’s effects on mediate its antidepressant anti-SI effects. was a secondary analysis of biomarker-focused, randomized, double-blind, crossover trial (0.5 mg/kg) compared saline placebo. Polysomnography (PSG) studies were conducted one day before after ketamine/placebo infusions. Sleep measured using spectral power functions over time including alpha (quiet wakefulness), beta (alert delta (deep sleep) power, as well macroarchitecture variables, wakefulness onset (WASO), total (TST), rapid eye movement (REM) latency, Post-Sleep Onset Efficiency (PSOSE). At baseline, diagnostic included lower TST ( p = 0.006) shorter REM latency 0.04) the versus HV group. Ketamine’s temporal dynamic (relative placebo) increased earlier night later night. However, there no significant patterns alpha, beta, or metrics, mediation variables These results highlight role sleep-related part systemic neurobiological initiated administration. Clinical Trials Identifier: NCT00088699.

Language: Английский

---

Increased connectivity of the anterior cingulate cortex is associated with the tendency to awakening during N2 sleep in patients with insomnia disorder

SLEEP, Journal Year: 2022, Volume and Issue: 46(3)

Published: Dec. 3, 2022

Abstract Study Objectives To investigate the relationship between sleep transition dynamics and stage-specific functional connectivity (FC) of anterior cingulate cortex (ACC) in patients with insomnia disorder (ID). Methods Simultaneous electroencephalography–functional magnetic resonance imaging (EEG–fMRI) data from 37 ID 30 well-matched healthy controls (HCs) were recorded during wakefulness different stages subsequently analyzed. A Markov chain model was used to estimate probability each stage. The FC ACC (set as seed) voxels across whole brain calculated. linear mixed effect determine group-by-stage interaction seed-based connectivity. correlation sleep-stage ACC-based explored. Results Patients exhibited a higher likelihood transitioning N2 than HCs. significant bilateral observed cerebellar, subcortical, cortical regions. Moreover, positive found cerebellum (r = 0.48). Conclusions This exploratory analysis indicates that enhanced represents potential neural pathway underlying greater waking sleep. These findings contribute an emerging framework reveals link maintenance difficulty function, further highlighting possibility is therapeutic target for meaningfully reducing disruption.

Language: Английский

---

A Sustainable Approach to Telerehabilitation in Europe: Patients Are Ready, but Caregivers Are Essential

Studies in health technology and informatics, Journal Year: 2024, Volume and Issue: unknown

Published: April 26, 2024

Background: Several studies have demonstrated the effectiveness of telerehabilitation. However, it remains unclear what proportion people in need rehabilitation can confidently use telecommunications networks and related devices. Objectives: The aim this study is to estimate patients who possess either requisite digital literacy perform telerehabilitation independently or a family caregiver capable providing effective support. Methods: Synthetic populations with realistic kinship network (i.e. trees) representative European countries are built. Age, sex, location-specific prevalence rates needs skills combined percentage digitally literate relatives. Results: In Europe, 86% potentially eligible for four out five cases, over age 65 require Conclusion: Telerehabilitation has potential spread Europe. Caregivers an essential social role ensuring sustainable access

Language: Английский

---