BioNorth Website

🌐 Live site: https://bionorthtech.github.io/BioNorth-Main/

✉️ Contact: biotech@bionorth.us

🏷️ Status: Phase 0 — Prototype Stabilization · Patent pending · LLC forming

---

Table of contents

• The problem
• Our solution
• How it works
• Technology
• The science
• Development roadmap
• Side projects
• The team
• Get involved
• Disclaimer

---

The problem

Freezing of gait (FOG) is a sudden, episodic inability to generate forward stepping. It affects over a million Americans with Parkinson's disease, progressive supranuclear palsy, and related conditions, and roughly 38% of Parkinson's patients report FOG episodes — which carry a 3× higher fall risk.

During an episode the patient intends to walk and the motor cortex sends the command, but abnormal beta-frequency synchronization in the basal ganglia intercepts the signal — a neurological lock between intention and execution.

Existing approaches fall short: medication is often dopamine-resistant for FOG; cueing requires manual activation and gives no physical assistance; deep brain stimulation is invasive; heel-switch FES triggers on biomechanics, not the brain; and prior BCI-FES systems trigger on motor intent alone — the same signal present during normal movement. None can distinguish a freeze from an intentional step.

Our solution

BioNorth detects the specific movement-failure state — motor intent present but muscle response absent — rather than triggering on intent alone. Stimulation fires only when a freeze is confirmed, and stops the moment EMG verifies the muscle has responded.

Key distinction (patent pending): every prior BCI-FES system stimulates when motor intent is detected. BioNorth stimulates when motor intent is detected AND muscle activation has failed to occur within a defined temporal window. This movement-failure criterion is the core patentable innovation.

How it works

1. Neural intent detected — EEG at Cz, FCz, Fz detects mu/beta desynchronization (movement intent).
2. EMG monitors response — electrodes on the tibialis anterior watch for muscle activation (300–500 ms window).
3. Failure state classified — intent present + no muscle response = a freeze.
4. Stimulation delivered — constant-current biphasic pulses drive dorsiflexion and break the freeze.
5. Confirmation & termination — EMG confirms activation; stimulation stops on verified response, not a timer.

Technology

Two wearable units — a headset EEG unit and a leg-band EMG + FES unit — communicating wirelessly, designed to run standalone with no phone or laptop required.

Subsystem	Now	Target
EEG acquisition	OpenBCI Ganglion (4ch, 200 Hz)	Custom ADS1299 PCB (8ch, 250 Hz)
EMG acquisition	Ganglion Ch4 differential (200 Hz)	Dedicated AD8221 InAmp (1 kHz)
Neurostimulation	NeuroStimDuino (2ch, I²C)	Custom MAX14521E FES circuit
Signal processing	Python via BrainFlow/LSL	nRF5340 dual-core firmware (C/Zephyr)

Signal chain: Motor Cortex → Scalp EEG (Cz/FCz/Fz) → ADS1299 ADC → nRF5340 DSP → Butterworth filters → band power (θ/μ/β) → state machine → [if freezing] → FES driver → constant-current pulses → tibialis anterior → EMG confirmation → [stim off].

The science

Every design decision traces to peer-reviewed movement-disorder research:

• Beta synchronization in the STN/motor cortex locks voluntary output during FOG despite intact intent → validates EEG beta ERD monitoring.
• Theta synchronization over frontal-central leads replaces beta ERD during FOG → basis for dual-band detection (μ/β ERD at Cz + θ ERS at Fz).
• Tibialis anterior is among the first muscles to cease at FOG onset → ideal EMG sentinel and stimulation target.
• Peroneal/TA FES for dorsiflexion has demonstrated safety and efficacy in foot drop → safety precedent and 510(k) predicate strategy.

Band	Range	Role
θ Theta	4–8 Hz	FOG onset marker
μ Mu	8–13 Hz	Motor idle rhythm
β Beta	13–30 Hz	Corticospinal drive
EMG	20–450 Hz	Muscle response confirmation

Development roadmap

A 28–36 month path from working prototype to FDA 510(k)-cleared device.

• P0 · Stabilize Prototype (in progress) — firmware fixes, electrode placement, movement-failure algorithm, baseline dataset, patent documentation.
• P1 · Validated Prototype — OpenBCI Cyton, MyoWare EMG + IMU, on-subject validation, provisional patent + PCT, first paper.
• P2 · Module Integration — ADS1299 + nRF5340 + MAX14521E stack, C/Zephyr firmware, custom PCB, mobile app, 3D-printed enclosures.
• P3 · Clinical Prototype — IRB pilot (10–20 subjects), ISO 14971 risk analysis, FDA pre-sub, utility patent, peer-reviewed publication.
• P4 · Regulatory — FDA 510(k), ISO 13485, manufacturing partner, NIH STTR Phase I.
• P5 · Commercial Launch — first production run, prescriber training, post-market surveillance, Series A, CE marking.

Side projects

Things we build between milestones:

• AiMetr — a multi-provider AI usage monitor with a real-time dashboard, a desktop-pet overlay, and an optional ESP32-S3 AMOLED hardware companion. Tracks rate limits, tokens, and cost across Claude, OpenAI/Codex, DeepSeek, Ollama, and LM Studio. MIT licensed.

More are on the way — watch the organization to catch them first.

The team

BioNorth is the work of a team of high-school students who came together to build it from the ground up — researching the neuroscience, wiring the hardware, and writing the detection algorithm in our own time. Along the way we've earned school awards and competed at science and engineering competitions. We're currently filing a provisional patent and forming an LLC to carry the work forward.

Get involved

We're seeking clinical research partners, biomedical-engineering mentors, and early-stage funding.

• 🤝 Clinical partners — neurologists, physiatrists, movement-disorder specialists for protocol co-development and IRB pilots.
• 🎓 Research mentors — biomedical-engineering faculty and researchers for signal-processing and clinical-validation collaboration.
• 💸 Early funding — angels, family offices, and patient-advocacy foundations.

📧 biotech@bionorth.us · 🐙 https://github.com/bionorthtech

Disclaimer

BioNorth is a research prototype. It is not FDA-cleared and is intended for investigational use only. Nothing here is medical advice.

---

© BioNorth Tech. All rights reserved. BioNorth™ and the BioNorth logo are trademarks of BioNorth Tech.

Tech stack

Vite + React (react-router-dom). Single-page app, deployed as a static site to GitHub Pages.

Develop

npm install
npm run dev      # local dev server
npm run build    # production build → dist/
npm run preview  # preview the production build

Project structure

src/
├── main.jsx            # React root + router
├── App.jsx             # routes + shared layout
├── theme.css           # design tokens, layout utilities, responsive rules
├── components/         # Nav, Footer, EEGWave, Reveal
├── pages/              # Home (the mission), SideProjects (AiMetr)
└── data/content.js     # site content constants

Deployment

Pushes to main trigger .github/workflows/deploy.yml, which builds the site and publishes it to GitHub Pages at https://bionorthtech.github.io/BioNorth-Main/.

One-time setup: in repo Settings → Pages, set Source = "GitHub Actions".

Custom domain (bionorth.us)

To serve from bionorth.us instead of the project path:

1. Set base: "/" in vite.config.js.
2. Add public/CNAME containing bionorth.us.
3. Configure the domain's DNS and the custom domain in Settings → Pages.