Somnison

Sleep compression technology

A wearable system that detects slow-wave sleep via dry EEG and delivers low-intensity focused ultrasound to enhance glymphatic clearance—potentially achieving full restorative sleep in less time.

Status
Proof of Concept
Mechanism
650 kHz FUS
Safety
10× Below FDA Limits
Core Hypothesis

Slow-wave sleep is not uniformly efficient. If we can mechanically amplify the natural CSF oscillation window, we can compress the sleep requirement.

During deep sleep, the brain's glymphatic system opens, allowing cerebrospinal fluid to wash through perivascular spaces and clear metabolic waste. This process is gated by slow oscillations—waves that occur approximately every 15–25 seconds, varying by individual.

Natural short-sleepers with DEC2 and NPSR1 mutations achieve full cognitive restoration in 4–6 hours. Their sleep is not just shorter—it is qualitatively more efficient at waste clearance per hour. Somnison attempts to mechanically induce this phenotype by enhancing glymphatic influx during the precise window when it naturally occurs.

60%
Interstitial expansion during SWS
~20s
Adaptive CSF oscillation period
6hr
Target vs. 8hr baseline
10×
Below FDA diagnostic limits

Closed-loop enhancement

The device does not blindly emit ultrasound. It listens for the precise neurophysiological signature of slow-wave sleep, then delivers calibrated mechanical stimulation timed to the user's natural CSF oscillation rhythm.

01

Detection

Four dry spring-pin EEG electrodes (Fp1, Fp2, F3, F4) continuously monitor delta-band power. A power-ratio classifier distinguishes SWS from lighter sleep stages. Impedance is checked in real time; if frontal electrodes fail, the system degrades gracefully to temporal channels.

Threshold: Delta/alpha-beta ratio > 2.5
Latency: < 30s classification window
Reference: Fz or linked mastoids (A1/A2)
02

Gating

Upon SWS onset, the system extracts the user's natural CSF oscillation period from the EEG envelope—typically 15–25 seconds. It predicts the next inflow window and fires a brief ultrasound burst during the low-resistance phase, when glymphatic channels are most open.

Carrier: 650 kHz
Envelope: adaptive period (user-specific)
Burst: ~1.5s ON, remainder OFF (~7.7% duty)
Target: MCA M1 segment, transtemporal window
03

Enhancement

Low-intensity focused ultrasound interacts with the natural arterial pulsation that drives glymphatic influx. The mechanism may involve acoustic radiation force adding directional momentum to perivascular fluid, or cellular mechanotransduction (TRPV4–AQP4) reducing hydraulic resistance to flow. Both effects amplify the existing clearance process without microbubbles or blood–brain barrier disruption.

Peak pressure: 0.2 MPa
MI: ~0.9 (FDA non-significant risk)
Delivery: bilateral temporal, 15° anterior, 10° superior
Coupling: integrated hydrogel pad (no user-applied gel)

Peer-reviewed foundation

Every claim is supported by published research. No animal studies using microbubbles are cited as primary evidence for the wearable protocol.

2013
Xie et al.
Science
Sleep Drives Metabolite Clearance from the Adult Brain

Demonstrated that natural sleep expands the brain's interstitial space by 60%, enabling convective exchange between CSF and interstitial fluid, and doubling β-amyloid clearance relative to wakefulness.

PubMed →
2019
Fultz et al.
NIH/PMC
Coupled Electrophysiological, Hemodynamic, and CSF Oscillations in Human Sleep

Identified large CSF oscillations at 0.05 Hz during NREM sleep in humans, coupled to delta slow-wave EEG activity—establishing the precise mechanical window for intervention.

PMC Full Text →
2025
Xiao et al.
Ultrasound in Med. & Biol.
Low-Intensity Transcranial Focused Ultrasound Enhances Glymphatic Transport

Non-microbubble FUS (650 kHz, 0.2 MPa, 7.7% duty cycle) significantly enhanced glymphatic transport across all anesthesia levels. Histologically safe at tenfold below FDA diagnostic limits.

PubMed →
2024
Airan et al.
Stanford / bioRxiv
Noninvasive Ultrasonic CSF Clearance in Brain Injury

Low-intensity transcranial FUS protocol drives CSF clearance via meningeal lymphatics without exogenous agents (no microbubbles, nanoparticles, or drugs). Effective awake, asleep, or semi-conscious. MI = 0.9.

PubMed →
2025
Lee et al.
JCBFM
Restoration of Glymphatic Influx After Stroke Using Low-Intensity Focused Ultrasound

Unilateral LIFU restored glymphatic CSF influx bilaterally within 10–20 minutes, peaking at 30–40 minutes post-stimulation—demonstrating rapid onset and non-localized effects.

PubMed →
2025
Advanced Science
Wiley
Very Low-Intensity Ultrasound Facilitates Glymphatic Influx via TRPV4-AQP4

Planar ultrasound at 3.68 mW/cm² enhanced glymphatic influx via TRPV4 mechanotransduction and AQP4 water channel opening—without microbubbles. Authors explicitly proposed wearable integration.

PMC Full Text →
2022
Dong et al.
iScience
Natural Short Sleepers Exhibit Accelerated Glymphatic Clearance

DEC2-P384R and NPSR1-Y206H mutations accelerated glymphatic clearance, significantly reducing amyloid and tau pathology despite 25–40% shorter sleep duration—proving efficient sleep is biologically possible.

PubMed →
2024
Qin et al.
Neurosci. & Biobehav. Rev.
Safety of Low-Intensity Transcranial Ultrasound Stimulation

Systematic review of 11 human RCTs and 44 animal studies confirmed therapeutic efficacy across multiple domains with no major adverse effects. Reversible minor symptoms resolved spontaneously.

PubMed →
2025
Liu et al.
J. NeuroEngineering Rehab.
Safety Review of Transcranial Focused Ultrasound (690 Participants)

Meta-analysis of 34 studies found zero major adverse events across 690 participants. Most common effects: scalp sensations (9.01%), sleepiness (7.17%), neck pain (5.93%). All resolved without intervention.

Springer →

Technical parameters

Prototype specifications for the Phase 0 proof-of-concept device. Subject to change during preclinical validation.

EEG Channels
4 (Fp1, Fp2, F3, F4)
EEG Electrodes
Dry spring-pin, gold-plated
Reference
Fz or linked mastoids (A1/A2)
Transducer Frequency
650 kHz
Transducer Placement
Bilateral temporal, transtemporal window
Beam Angle
15° anterior, 10° superior
Acoustic Coupling
Integrated hydrogel pad
Peak Pressure
0.2 MPa
Mechanical Index
0.9
Duty Cycle
~7.7% (adaptive)
Gating Period
Adaptive 15–25 s (user-specific)
Thermal Sensor
Contact thermistor, left transducer
Thermal Cutoff
< 38°C (hardware + software)
Audio
Bone conduction, bilateral temporal
Wake System
Amber LED sunrise simulation in mask
MCU
Teensy 4.1 (600 MHz)
ADC
ADS1299 (24-bit, 4-ch used)
Power
Split: left MCU / right battery
Classification Latency
< 30 seconds
Form Factor
Cap + integrated mask, adjustable band

Research updates

We are building the first closed-loop sleep compression device. If you are a sleep researcher, neurologist, or technical contributor, we would like to hear from you.

No consumer preorders. This is a research project.
MRG. Not a medical device.
Contact
research@somnison.com

Location
Lab address TBD

IRB Status
Not yet submitted

Webhook
Connected