Hey Neurohacker,

This is Weekly Signal #50.

That's 50 weeks of cutting through noise, dual-sourcing every story, and trying to give you the engineering lens on neurotechnology that I wanted to read but couldn't find anywhere. When I started this, I wasn't sure anyone else cared about the difference between a press release and a peer-reviewed result. Turns out, you do. So — genuinely — thank you for being here. Especially the 55% of you who open this thing every single week.

Now, a quick update before we get into this week's signals.

We launched hacktheneuron.com. It's live. There's a lot still to build — and I mean a lot — but with your support, it will get done. This project has always been community-first. That part doesn't change.

A few ways to keep the momentum going:

🧠 Refer a Neurohacker → If you know someone who thinks about brains the way we do — whether they're building, researching, or just deeply curious — send them our way. Refer a Neurohacker 🦾

✦ Check Out Our Sponsors → They make The NeuroCircuit possible. If their work is relevant to yours, show them some love. It directly supports what we're building here.

📬 Tell Us What You Want → This is the one I care about most. What do you want to see from us — more hardware deep dives? Policy breakdowns? Founder interviews? Community builds? Send Feedback This newsletter gets better when you shape it. 🧑‍💻

And before we get into this week's content — a quick word from our partner:

✦ This Week's Partner: Hubspot

Neurotechnology News | March 14–20, 2026

Let's cut through the noise.

This week, a BrainGate implant decoded bimanual finger typing at 22 words per minute in paralyzed users — the fastest intracortical BCI communication ever published — and the FDA granted its first Breakthrough Device Designation for a cognitive neural implant. Meanwhile, a second major law firm published a neural data compliance alert, signaling corporate legal teams are now actively pricing neurotech regulatory risk. Bottom line: BCI performance is closing in on able-bodied speed, and the compliance industry is mobilizing faster than the legislation itself.

BrainGate implant decodes finger-by-finger typing at 22 words per minute — the fastest intracortical BCI result ever published.

Two participants in the BrainGate2 trial — one with ALS, one with spinal cord injury — used intracortical microelectrode arrays to type on a virtual QWERTY keyboard by attempting individual finger movements. The faster participant hit 22 words per minute (110 characters/min) at a 1.6% word error rate. That approaches able-bodied smartphone texting speed. The system calibrated with as few as 30 sentences and operated in participants' homes — not a lab.

But the defining signal isn't the raw throughput. It's the architectural proof-of-concept underneath it.

The participant with 384 electrodes significantly outperformed the one with 128 — quantifying a 3x performance relationship between channel count and output fidelity. That bandwidth-performance curve now sets a concrete engineering target for every BCI team in the field.

It also encodes a tension: higher channel counts generate richer neural data that existing privacy frameworks aren't built to govern.

Published: March 16, 2026
Read more ➡️ Smithsonian / Nature Neuroscience

Two BrainGate2 participants used intracortical arrays to type on a standard QWERTY keyboard by attempting individual finger movements — achieving 22 WPM (110 char/min) with 1.6% word error. The highest-throughput intracortical BCI typing result published to date. The participant with 384 electrodes significantly outperformed the one with 128, and the system operated at home with as few as 30 calibration sentences.

💡 Big Picture: The 3x performance gap between 128 and 384 electrodes directly quantifies the bandwidth-performance curve in motor BCIs — every future improvement in output fidelity structurally requires more neural data, making the governance challenge harder at each step.

Published: March 18, 2026
Read more ➡️ Nature Neuroscience / Nature Neuroscience

Using a 3T MRI-compatible deep brain stimulation (DBS) system, researchers collected dense longitudinal neuroimaging data from Parkinson's patients across seven stimulation conditions over one year. DBS normalized the somato-cognitive action network (SCAN) but paradoxically denormalized effector motor circuits — the first evidence that current protocols involve an inherent circuit-level tradeoff.

Optimizing one network degrades another.

💡 Big Picture: This directly challenges open-loop DBS design and informs the control architecture of next-generation adaptive systems — ones that must tune stimulation parameters per-circuit, not globally.

Published: March 18, 2026
Read more ➡️ RTÉ News / Science Translational Medicine

Trinity College Dublin researchers used advanced MRI to image blood-brain barrier (BBB) integrity in 47 retired contact sport athletes (average 12 years post-retirement). A subgroup of 17 showed extensive persistent BBB leakage — correlated with worse cognition, reduced brain volume, and elevated monocyte counts. Cross-referencing with post-mortem chronic traumatic encephalopathy (CTE) tissue identified complement system activation (C5AR1, ITGAM, ITGB2, CD59) as a key pathological mechanism.

💡 Big Picture: BBB-focused MRI sequences could function as a non-invasive early-warning biomarker for CTE — a diagnostic tool that doesn't currently exist — while the complement pathway findings open a targetable pharmaceutical intervention window before irreversible neurodegeneration.

Published: March 18, 2026
Read more ➡️ University of Miami InventUM

University of Miami researchers targeted the mesencephalic locomotor region (MLR) — specifically the cuneiform nucleus in the brainstem — with DBS instead of conventional subthalamic or globus pallidus targets. All four Parkinson's patients showed improved mobility. A Michael J. Fox Foundation multisite trial is planned, alongside the first clinical trial of MLR-DBS for incomplete spinal cord injury..

💡 Big Picture: Shifting the DBS target from basal ganglia (indirect motor modulation) to a brainstem locomotor center (direct gait initiation) is a paradigm change in target selection — if validated, a single surgical approach could serve two distinct neurological conditions by amplifying residual descending pathways.

Published: March 19, 2026
Read more ➡️ Neuroscience News / Brain Stimulation

NYU Langone researchers demonstrated transcranial radio frequency stimulation (TRFS) using 945 MHz stub antennas that can bidirectionally modulate neural activity — either suppressing or exciting neurons — non-invasively in mice. A third stimulation modality, distinct from electromagnetic (TMS/tDCS) and ultrasound-based approaches.

Early-stage. Mouse data only. But the principle is new.

💡 Big Picture: TRFS adds a new entry to the non-invasive neuromodulation stack with different tissue-penetration and spatial-resolution tradeoffs — if it translates, it could address the persistent challenge of reaching deep brain structures non-invasively with focal precision.

Published: March 18, 2026
Read more ➡️ BioSpace / BusinessWire

Nia Therapeutics' Smart Neurostimulation System (SNS) earned FDA Breakthrough Device Designation for traumatic brain injury (TBI)-related memory loss — the first neurostimulation device to receive this designation for a cognitive indication. The fully implantable, wireless, closed-loop platform records from 60 channels across four brain regions, uses machine-learning classifiers to detect impaired memory encoding in real time, and delivered a 19% recall improvement in a sham-controlled study.

💡 Big Picture: This crosses the threshold from motor/sensory to cognitive restoration. The closed-loop architecture must decode a cognitive state (memory encoding failure), not just a pathological biomarker — creating a template that could extend neurostimulation into depression, PTSD, and other cognitive domains.

Published: March 20, 2026
Read more ➡️ IndexBox / Reuters via Yahoo Finance

Beijing-backed NeuCyber Neurotech publicly acknowledged its most advanced product — the Beinao-2 flexible-electrode invasive BCI — is roughly three years behind Neuralink, citing Neuralink's surgical robot and 20+ human patients as key advantages. NeuCyber has completed 7 human implantations of its earlier semi-invasive Beinao-1 and plans to expand to 50 patients this year, funded by ~200 million yuan (~$29M) in government backing.

💡 Big Picture: The candid gap assessment reveals two parallel strategies in China's BCI ecosystem: fast-track commercialization of simpler semi-invasive devices (as Neuracle demonstrated last week) alongside longer-term frontier development where surgical robotics and manufacturing remain the primary bottleneck.

Published: March 17, 2026 Read more ➡️ Cooley LLP / Morrison Foerster (update)

We covered the state-level neural data patchwork in Weekly Signal #48. This week, Morrison Foerster — one of the largest global law firms — published its own independent client alert cataloging legislative trends across seven states, identifying Illinois's private right of action (SB2994) and California's workplace neural monitoring restrictions (AB1883) as the highest-liability provisions.

When two top-tier firms issue overlapping client alerts within weeks, it means one thing: corporate legal departments are actively asking their outside counsel about neural data exposure. The compliance conversation has moved from academic to operational.

💡 Big Picture: The signal is no longer the legislation itself — it's that the compliance industry is mobilizing, which means neurotech companies without a neural data governance architecture will face pressure from investors, partners, and insurers before any of these bills are even enacted.

Can neural data governance scale at the same rate as electrode count?

This week's BrainGate data showed a 3x performance gap between 128 and 384 electrodes — meaning the pressure to increase channel count (and thus data sensitivity) is structurally engineered into BCI improvement.

At the same time, the FDA fast-tracked Nia's 60-channel cognitive implant with no federal neural data law in place, and corporate law firms are issuing client alerts to companies that haven't even started thinking about neural data governance.

The compliance industry is moving — but it's moving on a patchwork of unfinished state legislation with no federal floor. When the pressure to prepare arrives before the rules are finalized, companies have to design for uncertainty. And the question is whether "design for the most restrictive standard" is viable when no one yet knows what that standard will be.

Unlock a focused set of AI strategies built to streamline your work and maximize impact. This guide delivers the practical tactics and tools marketers need to start seeing results right away:

Stay ahead of the curve with these top strategies AI helped develop for marketers, built for real-world results.

Download the Free Report

Neurotech moves faster when people share what they're building. Whether you're working on a DIY biosignal project, following a clinical trial, or just had an idea you can't shake — put it out there. The best thinking in this field happens in the open.

Why bother? This community is small enough that your work actually gets seen, and focused enough that the people seeing it are the right ones. One post can get you a collaborator, a citation, or a conversation that changes the direction of a project.

The NeuroCircuit grows entirely through word of mouth. If this issue was worth your time, pass it to one person it would also be worth.

Share your referral link and earn:

Every issue gets shaped by what readers actually find useful. If something landed well, felt thin, or missed the mark entirely — say so. This is a small, focused publication and your feedback has a direct line to next week's issue.

What's worth a minute of your time: