Demo: NeuroStrip: A Direct‐to-Skin Wearable for Real-Time Neural Interface

We present NeuroStrip, a non-invasive, wearable neural interface that records high fidelity bioelectric signals (including surface electromyography, ECG, EOG, EEG style signals, and spatial motion data). At just ~4 g and as simple to apply as an adhesive bandage, it adheres directly to the skin to intercept physiological signals before visible motion occurs. It streams live in real time to the companion NeuroStrip app for visualization, recording, and synchronized video capture—enabling analysis of signal to action relations.

The NeuroStrip is a versatile neural interface platform that supports diverse use cases: • Sports science and performance tracking: Real-time muscle activation, biofeedback and analysis provides insights on over/under active muscles, asymmetry, recruitment and fatigue that guides customized treatment/performance plans
• Rehabilitation and clinical training: Leverage sEMG biofeedback for operant conditioning to accelerate recovery timelines, improve patient engagement and understanding • VR and AI driven interfaces: Bio signal based input/control systems for virtual environments, gesture decoding, or direct AI model integration • Multimodal bio signal capture: Seamless acquisition of sEMG, ECG, EOG, EEG-like and spatial signals for foundation model scale bio signal datasets

This demo aligns strongly with the workshop’s focus on wearable hardware, real-time decoding, foundation model enabling technologies, and bio signal analytics. We feel the device is superior to the present options in the market for the following reasons :-

• Ultra light direct-to-skin adhesive interface, that minimizes electrode-displacement, which in turn mitigates movement artifact delivering pre-movement signal fidelity and accuracy unseen before in other bio signal acquisition platforms. • Rapid application & training—devices can be applied in seconds, even by non specialists. • Structured protocol templates support reproducible data collection templates across users and sessions; enables cross user comparison, cohort studies, or teams. • Cloud based secure storage & authentication across workspaces and devices facilitates collaborative team access and multisite studies. • Enables training of AI models on high quality multimodal signals for generalized decoding across tasks and subjects. With structured data gathered at scale, the NeuroStrip in conjunction with the software interface could prove to be an invaluable tool for building bio signal foundation model development and more humanlike AI.

Setup involves applying the NeuroStrip adhesive patch to the skin, pairing with the NeuroStrip Controller app (on Windows/macOS/tablet) via Bluetooth, and selecting a structured protocol template within the app. The template launches synchronized video and bio signal capture, enabling live recording and post-session analysis.

The NeuroStrip wearable demonstrates a convergent platform for real-time bio signal decoding, structured data gathering, and AI model integration. We believe it exemplifies the cutting-edge wearable hardware and bio signal tools envisioned by the BrainBodyFM community and would be a compelling live demo at NeurIPS 2025.

Participants can stream and visualise live signals while performing simple gestures or facial muscle activation. They can view synchronised video and neural data side by side in structured sessions (pre-configured protocols) for a smooth recording experience. They can also see the recording recreated in post-session analysis for signal analysis and other applications like ML model training, AI analysis, outputs as PDF etc.