Wavefront is the first consumer health platform that uses sleep EEG brainwave data to predict migraine attacks up to 36 hours in advance — before you feel a thing.
88.8% of migraine sufferers want a device that can predict attacks. Today's tools can't. Patients live in constant anxiety about the next episode.
Current management is almost entirely reactive — take medication after the pain starts. Preventive treatments are broad, with no way to time them to actual risk.
Existing migraine apps rely on self-reported symptom diaries — subjective, inconsistent, and prone to recall bias. None use neurological data from the brain itself.
Research shows the brain's electrical activity shifts measurably up to 36 hours before a migraine — but no consumer product has been built to act on this signal. Until now.
Using consumer EEG sleep earbuds, Wavefront captures slow-wave sleep metrics every night. Disruptions in slow-wave activity are directly linked to cortical spreading depression (CSD), the neurological event underlying migraine with aura. This is the signal no other consumer product captures.
A smartwatch provides continuous autonomic nervous system data that captures pre-migraine physiological changes. HRV drops, resting heart rate shifts, respiratory rate changes, and wrist temperature fluctuations all serve as secondary confirmation signals.
A daily check-in captures prodromal symptoms, aura signals, triggers, stress levels, sleep quality, and alcohol intake. These subjective markers — especially neck stiffness, yawning, and light sensitivity — are known pre-migraine indicators.
Classification algorithms using EEG data consistently achieve 80–96% accuracy distinguishing migraine states, with deep learning architectures reaching 95.99% and neural networks hitting 97–98% precision across multiple studies.
Published research shows that combining EEG with biometric and behavioral signals significantly outperforms any single data source alone — validating Wavefront's multi-channel approach.
Prediction using only wearable biometrics and symptom diaries achieved a modest AUC of 0.62 — demonstrating that without EEG, prediction accuracy is fundamentally limited. This is the gap Wavefront fills.
Disrupted slow-wave sleep is directly linked to cortical spreading depression — the wave of neuronal depolarization that triggers migraine with aura. Monitoring slow waves provides a direct window into this process.
EEG data carries the dominant weight in the risk score. Biometrics and symptoms confirm and refine — but the brain signal leads.
The system begins with research-backed priors, then transitions to individualized weights as it learns each user's unique migraine patterns over time.
The algorithm doesn't just look at tonight — it analyzes multi-night patterns, trajectories, and contextual factors that single-night snapshots miss.
Multiple proprietary modifiers account for known migraine biology — including clustering behavior, refractory periods, and established trigger dynamics.
Over 50% of migraine sufferers lack proper medical management. The preventive treatment segment — where Wavefront plays — represents 62% of the market at $4.2B. Yet no existing solution uses objective brain data to time preventive interventions. The gap between what patients need (prediction) and what they have (reaction) is Wavefront's opportunity.
| Capability | Wavefront | Migraine Buddy | N1-Headache | Sonuby Weather |
|---|---|---|---|---|
| Sleep EEG brain data | ✓ Primary signal | ✗ | ✗ | ✗ |
| Predictive risk score | ✓ Nightly + tomorrow | ✗ Logging only | Limited | Weather only |
| Wearable biometric integration | ✓ Full smartwatch stack | Limited | ✓ Fitbit/Garmin | ✗ |
| Adaptive personalization | ✓ Data-driven weights | ✗ Static | Limited | ✗ |
| Multi-night pattern recognition | ✓ | ✗ | ✗ | ✗ |
| Behavioral trigger modeling | ✓ Research-grounded | ✗ | ✗ | ✗ |
| Barometric pressure (auto) | ✓ Geolocation API | ✗ | ✗ | ✓ Core feature |
| CSD tracking | ✓ Logged per night | ✗ | ✗ | ✗ |
Consumer EEG sleep earbuds provide clinical-grade slow-wave data during sleep. A smartwatch captures continuous biometric monitoring. Both are consumer devices requiring no clinical setting.
Beyond raw inputs, the algorithm computes proprietary night-over-night trends, multi-night trajectories, and composite ratios that capture patterns invisible in single-night data.
Working prototype with real patient data. Algorithm backtested with 100% sensitivity. Multi-channel data pipeline operational. Risk scoring, dashboards, and daily check-in live.
Onboard 10–20 migraine patients. Validate algorithm across diverse migraine patterns. Transition to data-driven adaptive weights. ML companion model for no-EEG nights.
Formal clinical study with neurologist partnership. Publish sensitivity/specificity data. Pursue FDA De Novo classification for digital migraine prediction software.
Server-side algorithm (IP protected). Physician dashboard for prescribers. Insurance reimbursement pathway. Integration with CGRP preventive medication timing.
We're seeking investment partners who understand the convergence of neurotechnology, data science, and personalized medicine.
Interested in learning more?