On Monitoring Brain Health from the Depths of Sleep: Feature Engineering and Machine Learning Insights for Digital Biomarker Development

McConnell, B. V.; Liu, Y.; Biswas, A. K.; Bettcher, B. M.; Medenblik, L. M.; Broussard, J. L.; Lucey, B. P.; Ramos, A. R.; Kheyfets, V. O.

doi:10.1101/2024.02.27.581950

bioRxiv · 10.1101/2024.02.27.581950

On Monitoring Brain Health from the Depths of Sleep: Feature Engineering and Machine Learning Insights for Digital Biomarker Development

McConnell, B. V.·Liu, Y.·Biswas, A. K.·Bettcher, B. M.·Medenblik, L. M.·Broussard, J. L.·Lucey, B. P.·Ramos, A. R.·Kheyfets, V. O.

View original source ↗Export metadata ↓

Abstract

BackgrounSingle-channel sleep electroencephalography (EEG) is a promising technology for creating cost-effective and widely accessible digital biomarkers for monitoring brain health. Sleep, notable for its numerous connections to brain health, is of particular interest in this context. Indeed, several of the best studied and widely recognized risk factors for neurodegenerative disease are also connected to aspects of sleep physiology, including biological sex, hypertension, diabetes, obesity/metabolic dysregulation, and immune system dysfunction. In this study, we utilize the unique signal characteristics of slow wave sleep (SWS) oscillatory events as features in machine learning models to predict underlying biological processes that are highly relevant to brain health. Our objective is to establish a foundation for algorithms capable of effectively monitoring physiological processes in sleep that directly and indirectly inform brain health using single-channel sleep EEG as a functional metric of brain activity. MethodsUtilizing data from the Cleveland Family Study, we analyzed 726 overnight polysomnography recordings to extract features from slow waves and adjacent oscillatory events. Advanced signal processing and machine learning techniques, including random forest models, were employed to engineer features and predict health-related outcomes such as age, cerebrovascular risk factors, endocrine functions, immune system activity, and sleep apnea. ResultsOur models demonstrated significant predictive capability for several outcomes, including age (R2 = 0.643, p < 0.001), and sex classification (area under the receiver operator characteristic (AUROC) curve = 0.808), diabetes and hypertension diagnosis (AUROC = 0.832 and 0.755, respectively). Significant predictions were also modeled for metabolic/endocrine functions (including blood concentrations of IGF-1, leptin, ghrelin, adiponectin, and glucose), and immune markers (including IL-6, TNF-alpha, and CRP). In addition, this approach provided successful predictions in regression modeling of BMI and both regression and classification of sleep apnea. DiscussionThis study demonstrates the potential of using features from oscillatory events in single-channel sleep EEG as digital biomarkers. These biomarkers can identify key health and demographic factors that both affect brain health and are indicative of core brain functions. By capturing the complex interactions of neural, metabolic, endocrine, and immune systems during sleep, our findings support the development of single-channel EEG as a practical tool for monitoring complex biological processes through metrics that originate in brain physiology. Future research should aim to refine these digital biomarkers for broader home-based applications that may utilize inexpensive "wearable" devices to provide a scalable and accessible tool for tracking brain health-related outcomes.

Explore related subjects

neuroscience Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

McConnell, B. V., Liu, Y., Biswas, A. K., Bettcher, B. M., Medenblik, L. M., Broussard, J. L., Lucey, B. P., Ramos, A. R., Kheyfets, V. O.. 2024-02-29. On Monitoring Brain Health from the Depths of Sleep: Feature Engineering and Machine Learning Insights for Digital Biomarker Development. https://doi.org/10.1101/2024.02.27.581950

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Connexin 40 deficiency alters the temporal profile of postictal oxygen dynamics following focal seizures.

Epilepsy is increasingly recognized as a disorder involving both neuronal and vascular dysfunction. While connexin signaling has been implicated in epileptogenesis, the contribution of vascular connexins to seizure associated cerebrovascular pathology remains poorly understood. Connexin40 (Cx40) is an endothelial gap junction protein that plays a crucial role in vascular communication and blood-flow regulation. Seizures induce dynamic changes in cerebral perfusion and oxygenation, including prolonged postictal hypoperfusion/hypoxia. To determine whether Cx40 influences postictal hypoxia following focal seizures, we examined seizure characteristics and postictal oxygen dynamics in Cx40 knockout (Cx40-/-) mice using an established focal hippocampal seizure model. Electrically kindled seizures were elicited in wild-type and Cx40-/- mice, and local hippocampal tissue oxygenation was continuously monitored before and after seizure induction. Seizure duration did not differ between genotypes, indicating comparable seizure severity. Interestingly, Cx40 deletion altered the temporal pattern of postictal oxygen recovery, producing greater early hypoxia and a delayed secondary rebound in pO2 despite similar peak oxygen levels and overall hypoxic burden. These findings demonstrate that loss of Cx40 selectively alters the temporal profile of postictal oxygen dynamics without affecting seizure duration. Taken together, the results suggest that endothelial gap junctional communication contributes to postictal vascular recovery and identify Cx40 as a potential modulator of seizure associated neurovascular dysfunction.

neuroscience↗

bioRxiv2026

Attention Across Scales: From Individual Variation to Social Hierarchies and Brain Networks in Semi-Free-Ranging Macaques

Attention is a fundamental brain function supporting perception, decision-making, and social behavior, and its dysfunction profoundly impairs daily life. It is both dynamic and stable, varying across observations and individuals, changing across the lifespan, and being shaped by social and environmental experience. Yet capturing this complexity remains a central challenge in neuroscience. Here, we integrated longitudinal behavioral assessments of semi-free-ranging macaques living in naturalistic social groups with resting-state fMRI. We quantified performance across days, ages, and social hierarchies and related it to intrinsic brain organization. Distinct attentional phenotypes emerged, including individuals with reduced attentional control. Performance followed an inverted-U lifespan trajectory, improving from childhood to adulthood before declining. Social status modulated attentional performance. Critically, nonlinear lifespan trajectories and associations with individual attentional differences were most clearly expressed in frontoparietal connectivity. Together, these findings reveal how sustained attention is organized across scales, providing a biological framework for its individual diversity, social modulation, and neural basis.

neuroscience↗

bioRxiv2026

Decoding natural scenes from patterned optogenetic responses in mouse visual cortex

A central challenge in developing visual cortical prostheses is to determine how visual stimuli should be transformed into effective patterns of cortical stimulation. Although advances in stimulation technologies, including optogenetics, provide increasingly precise control over cortical activity, it remains unclear whether artificially evoked activity can reproduce the information content of naturally evoked visual representations. Here we establish a quantitative framework for evaluating visual encoding strategies by decoding cortical responses evoked by natural vision and patterned optogenetic stimulation. We developed a novel dual-modal paradigm in awake mice to bridge the gap between endogenous photostimulation and artificial network driving. By co-expressing the high-performance calcium indicator GCaMP6s and the red-shifted, ultra-sensitive opsin rsChRmine-oScarlet in the primary visual cortex (V1), we successfully translated dynamic natural movie frames into patterned, spatiotemporal optogenetic stimulation. Quantitative comparisons of macro-scale dynamics demonstrated that this patterned optogenetic injection evokes cortical states highly comparable and representationally aligned with those driven by actual visual photostimulation. To systematically evaluate the fidelity of these responses, we developed STAR, a deep learning model featuring spatial and temporal attention mechanisms, and successfully reconstructed the frames of natural movies from V1 signals under both experimental modalities. Collectively, our results demonstrate that complex sensory information can be both naturally encoded and synthetically injected into V1 circuits with high decoding fidelity. This work provides an empirical and computational proof-of-concept for intelligent, closed-loop biomimetic encoders, establishing a robust framework for next-generation cortical visual neuroprostheses and bidirectional brain-machine interfaces.

neuroscience↗