EMG Sensing Through Everyday Eyewear

Seeing the Mind
Through the Eyes

BlinkAid embeds EMG sensing electrodes into ordinary eyeglass nose pads to continuously monitor eye and facial muscle activity — capturing signals that cameras and EEG cannot.

BlinkAid NeuroClips clip-on prototype mounted on standard eyeglasses
62M+
Patients with PD or epilepsy worldwide
$170B
Combined global disease burden
0
Comfortable daily monitoring tools exist
The Signal

This is what our sensor sees

EMG electrodes in the nose pads capture distinct electrical signatures for every eye and facial movement — in real time.

Frown facial expression Real-time EMG signal during frown
Frown
Gaze left facial expression
Real-time EMG signal during leftward gaze
Gaze Left
Blink facial expression Real-time EMG signal during blink
Blink
Gaze right facial expression
Real-time EMG signal during rightward gaze
Gaze Right

These are just four of many detectable facial muscle activations. BlinkAid can decode the full range of periocular Action Units.

🧬 Facial Action Coding System (FACS)
Explore the full atlas of Action Units →
Hardware

18 months. Three prototypes. Real data.

We build, test, and iterate. Every prototype produced real EMG data from real people.

April 2023

Dev Prototype

First functional proof of concept. Exposed wires, breadboard electronics. Proved EMG signal capture from nose-pad electrodes is feasible.

Sep 2024

“PulseTech” Frame

Integrated glasses frame with custom PCB. First wearable form factor. 8-channel EMG + accelerometer. Used in driving simulator experiments.

Sep 2025

“NeuroClips” Clip-on

Universal clip-on module for any existing glasses. Smaller, lighter, discreet. Demonstrated at SILMO 2025 optics expo.

✓ Demonstrated — Our Data

Drowsiness & Fatigue Detection

Our first validated application. Real device, real subjects, real EMG data. BlinkAid detects physiological muscle fatigue before visible signs appear.

Drowsiness detection driving simulator experiment
BlinkAid EMG annotation dashboard with real-time signal traces and driving simulator video
Power spectral density analysis across sessions

What we measured

EMG power spectral density shifts reliably from 30–100 Hz (alert) to 10–30 Hz (fatigued). Combined with blink rate and blink duration changes, our sensor distinguishes alert from drowsy states in simulated driving experiments.

The Science

Why EMG-based eye monitoring works

Four decades of neuroscience research establish that eye behavior is a window into brain function. BlinkAid's EMG approach captures signals that cameras and EEG cannot.

Dopamine

Blink rate reflects dopaminergic tone

The basal ganglia regulate blinking through a GABAergic circuit involving the substantia nigra. Reduced dopamine in PD directly reduces spontaneous blink rate. L-dopa administration restores it.

Seizure

Seizures produce detectable eye EMG signatures

Eyelid myoclonia generates high-amplitude, rhythmic EMG bursts at 3–6 Hz. Ictal blinking produces stereotyped orbicularis oculi activation. These are directly measurable via periocular EMG electrodes.

EMG Advantage

EMG captures what cameras miss

EMG senses muscle activation directly — including sub-visible micro-contractions, eyelid dynamics during closure, and muscle fatigue patterns. It works in any lighting, any head position, and through closed eyelids.

Digital Biomarkers

Rich multi-dimensional feature space

Beyond simple blink counting: blink duration, amplitude, symmetry, inter-phase pauses, saccade latency, micro-saccade frequency, and eyelid velocity all carry diagnostic information.

🔭 Research Directions — Published Evidence

Where the platform goes next

Peer-reviewed research validates the feasibility of EMG-based monitoring for neurological conditions. These are research directions we are actively exploring — supported by decades of published clinical evidence.

Next-generation BlinkAid prototype AI-generated concept
Epilepsy

Seizure Detection

Published research shows that eye-based signals — blink patterns, gaze changes, and periocular EMG — can detect seizure events with clinical-grade accuracy.

88.9% Seizure Detection from Eye Tracking

A glasses-type eye tracker achieved 88.9% sensitivity for absence seizure detection with zero false alarms and detection latency of only 2.87 seconds.

Matsumoto et al., Clinical Neurophysiology 2021

Wearable EMG Detects 93.8% of Tonic-Clonic Seizures

In a multicenter study of 71 patients (3,735 hours of monitoring), a single-channel wearable EMG device detected generalized tonic-clonic seizures with 93.8% sensitivity and 9-second median latency.

Beniczky et al., Neurology 2018
50M+
People with active epilepsy worldwide (WHO)
$119B
Global annual cost of epilepsy (Begley et al., 2022)

Illustrative: Normal Blinking vs Eyelid Myoclonia

Normal blink (~0.3 Hz) Eyelid myoclonia (~4 Hz burst)
Conceptual illustration based on published literature — not BlinkAid data

Beniczky et al., Neurology 2018 — Demonstrated on a single-channel bicep-worn EMG sensor. BlinkAid’s 8-channel periocular EMG targets comparable or superior performance.

Parkinson’s Disease

Continuous Monitoring via Blink Biomarkers

Blink rate is a direct, non-invasive readout of central dopaminergic tone. Published research shows that continuous blink monitoring can track medication response and disease progression.

Reduced Blink Rate = Dopamine Deficit

Healthy adults blink 15–24 times/min. PD patients drop to 3–15/min. This is directly correlated with striatal dopamine levels, confirmed by neuroimaging.

Karson, Neurology 1984 · Vasudevan et al., Scientific Reports 2025

Blink Dynamics Track Medication State

ML models using blink features achieved AUCROC of 0.87 for ON/OFF state classification — outperforming even plasma drug levels for dyskinesia detection (0.86 vs 0.45).

Nishikawa et al., npj Parkinson's Disease 2025

Wearable Eye-Tracking Already Validated

Iwaki et al. used electrooculography eyeglasses in PD patients and achieved AUC of 0.902 for detecting wearing-off states — directly comparable to plasma L-dopa measurements.

Iwaki et al., Internal Medicine 2019
11.8M
People with Parkinson's disease worldwide (2021), projected to reach 25.2M by 2050
$51.9B
Annual economic burden of PD in the US alone (Michael J. Fox Foundation)

Illustrative: Healthy vs Parkinson's Blink Rate

Healthy (~18 blinks/min) Parkinson's (~7 blinks/min)
Conceptual illustration based on published literature — not BlinkAid data
Landscape

How BlinkAid compares

Approach Comfortable Daily Wear Works Anywhere EMG Dynamics Continuous PD Biomarkers Seizure Detection
BlinkAid (EMG Eyewear) ✓ ✓ ✓ ✓ ✓ ✓
Camera Eye Tracking ✗ ✗ ✗ ✗ ✓ ✓
Clinical EEG ✗ ✗ ✗ ✗ ✗ ✓
Adhesive EOG ✗ ✓ ✗ ✓ ✓ ✓
Wrist Wearables (Accel/HR) ✓ ✓ ✗ ✓ ✗ Motor only
BlinkLab (Camera) ✗ ✗ ✗ ✗ ✓ ✗
Team

BlinkAid’s Leadership

Yechiam Sapir
CEO

Yechiam Sapir

PhD, Neuroscience

Deep experience in medical device innovation. Former Head of Innovation at Surgical Monitoring. PhD in Neuroscience from Hebrew University, Harvard postdoctoral fellowship.

Raz Karl
CTO

Raz Karl

B.Sc Computer Science, B.Des Industrial Design

Over a decade in software, cybersecurity, and product design. Dual degrees from Hebrew University and Bezalel. Award-winning FIRST robotics team leader.

Iftach Yassur
CMO

Iftach Yassur

MD, Ophthalmology

Senior ophthalmologist and head of Oculoplastics at Rabin Medical Center. Over 25 years of clinical experience. Leading expert in blink disorders and eyelid surgery.

The future of neurological monitoring
is sitting on your nose

BlinkAid is building the first comfortable, continuous, EMG-based neuro-sensing platform for everyday eyewear. We’re looking for clinical partners, investors, and collaborators.

Contact Us