A small wireless patch captured water-sensitive brain signals through natural sleep, opening a potential path to home-based research into how sleep stages shape brain-fluid dynamics.
Study: A soft wearable near-infrared spectroscopy system for detecting brain water dynamics linked to glymphatic activity during sleep. Image Credit: Inkoly / Shutterstock
In a recent study published in the journal Science Advances, researchers introduced a soft, wireless, all-in-one, skin-conformal near-infrared spectroscopy (NIRS) sensor. Designed as a small, non-invasive forehead patch, the device was evaluated during 16 overnight home sleep recordings to track water-sensitive optical signals across different sleep stages.
This device offers a wearable alternative for monitoring brain-water-related signals outside restrictive laboratory environments, demonstrating, to the authors’ knowledge, that such dynamics can be recorded non-invasively during natural sleep at home.
Background
Decades of research on the physiological implications of sleep have overturned the historical belief that sleep is merely a passive state of rest, revealing it instead to comprise an active period of neural recovery, memory consolidation, and metabolic maintenance.
This metabolic maintenance, in particular, is associated with the relatively recently characterized glymphatic system. This network supports cerebrospinal fluid (CSF) circulation and the clearance of cellular waste products that accumulate during wakefulness.
Previous neurobiological investigations in highly controlled clinical environments have found that sleep disruption is associated with altered CSF dynamics and the accumulation of metabolic waste products, including amyloid-beta peptides implicated in Alzheimer’s disease (AD), although the causal relationships between sleep, glymphatic clearance, and neurodegeneration remain under investigation.
Unfortunately, current methods for investigating CSF and glymphatic-related processes, especially those that do so in real time, rely on invasive procedures or highly restrictive imaging, such as magnetic resonance imaging (MRI). Polysomnography (PSG), meanwhile, remains the clinical standard for characterizing sleep but is bulky and difficult to use for repeated home monitoring.
These approaches are known to be costly, technically demanding, and poorly suited to prolonged natural sleep measurements, necessitating the development of a novel system capable of non-invasively capturing brain-water-related optical dynamics within a naturalistic home environment.
About the System
This study aimed to address this requirement by designing a novel wireless, skin-conformal NIRS patch. The device featured three main layers: 1. A protective encapsulation layer composed predominantly of silicone, 2. A flexible printed circuit board (fPCB), and 3. A biocompatible skin adhesive.
The integrated fPCB contains multi-wavelength light-emitting diodes (LEDs) operating at 640, 680, and 950 nanometers, along with a specialized photodetector to capture their reflections. The system’s novel design functions by measuring near-infrared light reflected back through the user’s forehead tissues, thereby allowing it to estimate changes in oxygenated hemoglobin, deoxygenated hemoglobin, and water-related signals using the modified Beer-Lambert law.
The system’s performance was first assessed using Monte Carlo simulations of photon propagation through layered head tissues. These simulations showed that superficial tissues contributed most strongly to the signal, although a measurable fraction of detected photons was predicted to sample cortical gray matter beneath the scalp, the skull, and the CSF layer.
The patch’s safety and mechanical properties were subsequently tested, followed by a final round of in vivo human feasibility testing involving four healthy male participants aged 28 to 37 years. The study’s methodology required participants to wear the forehead patch during 16 overnight sleep sessions in participants’ own homes.
To establish reference sleep-stage measurements, the NIRS patch data were continuously compared with synchronized electroencephalogram (EEG) and electrooculogram (EOG) data gathered from a commercial sleep monitoring system.
Study Findings
The study found that the flexible fPCB device demonstrated consistent performance and signal stability under a range of physical conditions. The patch was found to achieve an average signal-to-noise ratio (SNR) of 12.72 decibels (dB) while sitting and 9.5 dB while climbing, with higher values than a rigid PCB implementation across sitting, walking, and climbing tests.
When evaluated using a hybrid artificial intelligence model that combined supervised machine learning (ML) algorithms with sigma-band neural threshold pipelines, concurrently recorded EEG and EOG signals were classified into wake, non-rapid eye movement (NREM), and rapid eye movement (REM) stages with approximately 80-90% accuracy. The NIRS-derived signals were then aligned with these reference sleep stages rather than being used independently to classify sleep.
NIRS-derived brain water signal data, in particular, exhibited tightly coupled, state-dependent shifts that mirrored the neural architecture of typical human sleep. Spectral analyses of the shifts identified peaks within frequency ranges commonly associated with respiration (~0.3 Hz), the cardiac cycle (~0.8 to 1.2 Hz), and slow oscillation-linked NIRS oscillations (SONO; 0.6 to 0.7 Hz). However, these physiological assignments were not quantitatively validated against simultaneous respiratory, cardiac, or gold-standard polysomnographic reference signals.
Group-level analysis demonstrated that entering NREM sleep was associated with increases in the cumulative NIRS-derived water signal, whereas transitioning from NREM to REM was associated with decreases in this signal. These accumulated traces described the direction and persistence of changes in Δ[H2O] and should not be interpreted as direct measurements of absolute brain-water accumulation or CSF flow.
Conclusions
This study is likely the first to successfully demonstrate that a soft, wearable NIRS device can continuously monitor sleep-linked, water-sensitive optical dynamics in a naturalistic home setting. This technology may support future longitudinal research into sleep physiology and brain-fluid dynamics, potentially associated with glymphatic activity.
While the authors do acknowledge the need for future iterations to incorporate short-separation channels to better account for superficial scalp contributions, this unobtrusive patch could eventually complement established research tools for studying neurological conditions in which sleep and brain-fluid regulation are altered. However, the device did not directly measure CSF flow, waste clearance, glymphatic function, or disease biomarkers, and its diagnostic value remains untested in patients.
Larger studies involving more diverse populations and simultaneous reference measurements, such as MRI or disease-specific biomarkers, will be required before clinical applications can be considered.
Journal reference:
- Ban, S., et al. (2026). A soft wearable near-infrared spectroscopy system for detecting brain water dynamics linked to glymphatic activity during sleep. Science Advances, 12(28). DOI: 10.1126/sciadv.aed2056. https://www.science.org/doi/10.1126/sciadv.aed2056
Researchers identify brain circuits responsible for triggering OCD symptoms