NeuralWear

Our NeuralWear1.0 takes the form factor of an ear-worn wearable that wraps around the ears like a hearing aid. It can capture important head-based biosignals, including.

• Brain waves (Electroencephalogram, EEG).

• Eye movements (Electrooculography, EOG).

• Facial muscle contractions (Electromyography, EMG).

• Galvanic skin responses (GSR).

• Heart rate and heart rate variability (HRV).

• Motions (Accelerations, orientation and angular velocity).

Our NeuralWear1.0's biosensing earpieces (EEG/EOG/EMG) have been proven robust to environmental noise and motion artefacts, thanks to our novel designs featuring active electrodes and multi-stage pre-amplification directly at the sensing sources:

• Ultra-high input impedance (10TΩ): 1000-10000X higher than state-of-the-art EEG recorders (e.g. ANTNeuro, OpenBCI) or conventional medical standard (>100MΩ). This ensures the highest signal fidelity and minimal coupled electromagnetic noise.

• 100X on-electrode pre-amplification: tiny uV biosignals are amplified directly on the electrode by using a novel feed-forward differential amplification method. The output biosignals are 100X amplified and converted to differential mode to ensure robustness over transmission wires.

Our NeuralWear1.0 wirelessly streams captured biosignals to a computer or mobile device over Bluetooth for real-time visualisation and processing. We provide a customised driver for our NeuralWear to work directly with the OpenVIBE framework.

Our NeuralWear1.0 features a tinyML framework and a neural accelerator to efficiently run embedded AI models directly on the device. The NeuralWear1.0 adopts Arduino NICLA Voice as the main controller unit with an on-board Syntiant NDP120 AI accelerator.