The Science

The research behind brain-sensing earbuds

Most wellness technology asks you to take its word for it. We would rather show our work. Below is the peer-reviewed science — both the studies NextSense has run and the foundational neuroscience the technology stands on. Published, not promised.

Our research

Studies NextSense has run or contributed to — the clinical validation that takes brain sensing out of the lab and into the ear.

Brain-Guided Audio: Decoding Cognitive and Physiological States from Ear-EEG, from Sleep to Focus

Berent (NextSense) · NextSense Technical White Paper, 2026

A NextSense review of decoding a hierarchy of states from in- and around-ear EEG — sleep, alertness, attention, and calm — and the brain-guided audio framework in which decoded brain states drive personalized sound. Argues ear-EEG is already ready for sleep and drowsiness, and promising but not yet proven for focus and calm neurofeedback. Published here in full.

Beyond Band Power: Aperiodic (1/f) and Waveform-Aware Analysis of Ear-EEG

Berent (NextSense) · NextSense working paper (target: Journal of Neural Engineering), 2026

On real NextSense ear-EEG, naive band power can invert conclusions. Using a parameterized spectral model (specparam) and cycle-by-cycle metrics across two cohorts, the aperiodic 1/f exponent steepens monotonically into deep sleep (Δ N3−WAKE = +1.05, N3 steepest in all 15 subjects), while aperiodic-aware features decode sleep as well as band power — with a montage dependence that is itself a methodological result.

Real-Time Multimodal Sleep Staging from Consumer Wearable Sensors Validated Against Ear-EEG: A Study Protocol

Berent (NextSense) · NextSense study protocol (target: JMIR Research Protocols / npj Digital Medicine), 2026

A pre-registered study protocol to test whether a consumer smartwatch (heart rate, accelerometer, microphone) can stage sleep in real time, validated against simultaneously recorded ear-EEG as the reference standard (Cohen’s κ non-inferiority to 0.61). Pre-data; execution pending IRB approval.

Toward Focus Decoding from Ear-EEG: A Shipped Measurement Pipeline and a Corpus-Collection-at-Scale Design

Berent (NextSense) · NextSense Technical White Paper, 2026

A measurement-first white paper: a shipped ear-EEG pipeline that logs two grounded markers — relative alpha and the aperiodic 1/f exponent — under artifact gating to build a labeled, real-world corpus, and deliberately ships no focus score and no closed loop until the data warrant one.

Using a standalone ear-EEG device for focal-onset seizure detection

Joyner, Hsu, Martin, Dwyer, Chen, Sameni, Waters, Borodin, Clifford, Levey, Hixson, Winkel & Berent (NextSense, Emory) · Bioelectronic Medicine, 2024

In-ear EEG validated against intracranial EEG: 86.4% of focal seizures detected across 1,255+ hours in 20 patients, with a false-alarm rate of 0.1/day. The first study to systematically compare in-ear EEG against simultaneous intracranial EEG — establishing that clinical-grade brain sensing can leave the lab and live in the ear.

A novel, wearable, in-ear EEG technology to assess sleep and daytime sleepiness

Berent, Saini, Stewart, Bheda, Borodin, Paul, Tracey, Dong, Volfson, Buhl & Rye (NextSense) · Bioelectronic Medicine, 2026

Extends in-ear EEG from seizure detection to sleep — assessing sleep architecture and daytime sleepiness from earbuds, the foundation for reading and improving sleep at home.

Wireless In-Ear EEG Denoising and Virtual Channel Modeling for Sleep Staging

Velez, X. (Georgia Tech) — data from the Saini group at Emory University with NextSense · Research collaboration, 2025

After movement-artifact removal and deep-learning signal modeling, in-ear EEG captured 92.7% of the sleep-specific information found in the gold-standard scalp electrode — with particular strength at deep (N2/N3) sleep, the stages that matter most for restorative, brain-clearing slow-wave sleep. Public link forthcoming.

Publication forthcoming

The ear as actuator: neuromodulation

The ear is not only the best place to read the nervous system — it is a place to act on it. Electrical stimulation at the ear’s vagus nerve (Nēsos), and acoustic and sensory entrainment that nudges the brain toward a target rhythm, both produce real, measured effects — on inflammation, mood, alertness, and even the pathology of Alzheimer’s. The therapeutic counterpart to sensing, much of it in the same real estate.

Non-invasive vagus nerve stimulation for rheumatoid arthritis: a proof-of-concept study

Nēsos — proof-of-concept clinical trial · The Lancet Rheumatology, 2021

A wearable, non-invasive device delivering stimulation to the auricular branch of the vagus nerve — through the ear — reduced disease activity in rheumatoid arthritis, with no serious adverse events. Proof that the ear is not only a place to read the nervous system, but a place to act on it.

Transcutaneous auricular vagus nerve stimulation for major depressive disorder with peripartum onset (DELOS-1)

Nēsos / NextSense — DELOS-1 trial · Journal of Affective Disorders, 2022

A multicenter trial of a wearable auricular vagus-nerve-stimulation system for peripartum depression, reporting strong response and remission with a non-invasive, non-pharmacological treatment — the therapeutic counterpart to brain sensing, delivered through the same real estate: the ear.

Auditive beta stimulation as a countermeasure against driver fatigue

Moessinger, Stürmer & Mühlensiep (Renault / Infrasonics) · PLoS ONE, 2021

Sound tuned to a ~18 Hz beat, delivered through the ears, measurably increased EEG beta and reduced drowsy theta in 80 drivers — cutting subjective fatigue and speeding reaction times for over an hour. Acoustic entrainment that shifts brain state, confirmed on EEG.

Innovations in noninvasive sensory stimulation treatments to combat Alzheimer’s disease

Park & Tsai (MIT Picower) · PLoS Biology, 2025

40 Hz light and sound (GENUS) drive gamma brain rhythms that, in models, clear amyloid and tau, mobilize the brain’s cleanup cells, and improve memory — with auditory 40 Hz alone reaching deep memory structures. Early human evidence is ongoing. The clearest sign that the right rhythm, delivered through the senses, changes the brain’s biology.

The science we’re built on

We did not invent the brain. These are the landmark findings — on attention, the aperiodic signal, deep sleep, and the measurable biology of belief — that the technology is designed around.

Ear-EEG Devices for the Assessment of Brain Activity: A Review

Juez, Moumane, Nassar, Molina-Salcedo, Segura-Quijano, Valderrama & Le Van Quyen · IEEE Sensors Journal, 2024

A systematic review of 96 peer-reviewed ear-EEG studies since 2011, across sleep, epilepsy, brain-computer interfaces, and more. The evidence that ear-EEG is an established field — and that the ear, beneath the temporal lobe and wired to the vagus nerve, is privileged real estate for reading the brain.

From Scalp to Ear-EEG: A Generalisable Transfer Learning Model for Automatic Sleep Scoring in Older People

Hammour, Davies, Atzori, della Monica, Ravindran, Revell, Dijk & Mandic · IEEE Journal of Translational Engineering in Health and Medicine, 2024

A sleep-scoring model trained only on scalp EEG, applied to a single in-ear sensor in older adults, scored sleep at 70.1% out of the box and 73.7% after light fine-tuning — with the biggest gains on deep (N3) sleep. Evidence that decades of scalp knowledge transfer to the ear.

Preparatory encoding of the fine scale of human spatial attention

Voytek, Samaha, Rolle, Greenberg, Gill, Porat, Kader, Rahman, Malzyner & Gazzaley · Journal of Cognitive Neuroscience, 2017

Preparatory alpha (8–12 Hz) activity encoded where attention was directed and predicted accuracy and reaction time nearly a full second before a target appeared. Alpha is the brain aiming attention — not merely the frequency of calm.

Behavioral and cognitive correlates of the aperiodic (1/f-like) exponent of the EEG power spectrum

Ostlund, Alperin, Drew & Karalunas · Developmental Cognitive Neuroscience, 2021

The popular theta/beta ratio is confounded by the brain’s aperiodic (1/f) background. The real signal is the aperiodic exponent, which tracks excitation/inhibition balance — a reminder that honest measurement means separating signal from background.

Placebos without Deception: A Randomized Controlled Trial in Irritable Bowel Syndrome

Kaptchuk, Friedlander, Kelley, Sanchez, Kokkotou, Singer, Kowalczykowski, Miller, Kirsch & Lembo · PLoS ONE, 2010

Patients openly told they were taking inert placebo pills still improved significantly more than untreated controls. Expectation produces real change — which is exactly why self-report alone cannot tell genuine effects from belief.

Justice for Placebo: Placebo Effect in Clinical Trials and Everyday Practice

Knezevic, Sic, Worobey & Knezevic · Medicines, 2025

A review of the neurobiology of placebo: dopamine, endogenous opioids, and endocannabinoids, visible on PET and fMRI, with effects reaching objective markers like inflammation. Belief leaves measurable traces — so measurement is the honest referee.

Acoustic Enhancement of Sleep Slow Oscillations

Papalambros et al. · Frontiers in Human Neuroscience, 2017

Pink-noise pulses timed to the slow-wave upstate increased slow-wave activity and memory — the closed-loop principle behind reading the brain in real time and responding within the sleeping rhythm.

Sleep Drives Metabolite Clearance from the Adult Brain

Xie et al. · Science, 2013

During deep sleep, the brain’s glymphatic system accelerates clearance of metabolic waste, including amyloid beta. The reason deep, slow-wave sleep is worth measuring — and protecting.

Real science, not wellness hype

See the clinical evidence behind NextSense

Peer-reviewed studies and hospital-grade validation with collaborators at Emory, Northwestern, McGill, and Mayo — the science behind brain-sensing earbuds that read your rhythm and respond.

Explore NextSense