Devices

Paradromics’ Connexus intracortical BCI platform, validated in a short intraoperative first-in-human recording and positioned to scale into a chronic, fully implantable communication/speech system.

Device — Intracortical

Paradromics Connexus (acute first-in-human recording)

Official site →

BCI · intracortical · Paradromics · Connexus · acute · intraoperative · speech restoration · cortex · recording · array

Paradromics Connexus (acute first-in-human recording)

One-line verdict: A high-channel-count intracortical implant platform validated in a short intraoperative human recording, designed to scale into chronic, fully implantable speech/communication BCIs.

Quick tags: Recording · Intracortical · Acute intraoperative validation · Species: Human · First in-human recording: 2025

Overview

What it is: Connexus is Paradromics’ intracortical BCI implant platform. In a first-in-human milestone, the device was temporarily implanted during an epilepsy resection surgery to verify that it can be safely implanted, record human neural signals, and be removed intact in a short intraoperative window.

Why it matters: This is a canonical example of the modern “acute intraoperative validation → chronic feasibility study” pathway for high-density intracortical BCIs: de-risk the mechanics of implantation/explant and basic signal acquisition before committing to chronic implants.

Most comparable devices: Neuralink N1 (intracortical), BrainGate Utah-array systems (intracortical), high-density surface systems (contrast).

Spec Card Grid

Identity

  • Device name: Connexus
  • Canonical ID: BTSD-IMBCI-0010
  • Org / manufacturer: Paradromics (Austin, TX, USA)
  • First demonstrated (year): 2025 (first in-human recording announced)
  • First implanted (year): 2025 (acute intraoperative placement)
  • Species: human (acute intraoperative)
  • Regulatory / trial status: acute intraoperative test reported (2025); chronic feasibility efforts/IDE announcements reported subsequently
  • Primary use: recording
  • Primary target: cortex (case-specific; intraoperative epilepsy surgery context)

Geometry & Architecture

  • Interface type: intracortical microelectrode array (penetrating)
  • Penetrating?: yes
  • Form factor: “cortical module” (described publicly as smaller than a dime)
  • Array layout: dense multi-electrode implant (exact geometry not fully specified in public sources)
  • Footprint (mm): ~dime-scale module (reported)
  • Insertion depth (mm): ~1–2 mm class reported in secondary sources for similar implants (treat as estimate until a primary technical source is cited)
  • Shank / lead dimensions: not publicly standardized in a peer-reviewed technical spec
  • Site spacing (µm): not publicly standardized
  • Tip geometry: microelectrode tips (intracortical)
  • Insertion method: surgical intracortical placement (intraoperative)
  • Anchoring method: system-dependent
  • Packaging location: marketed as a fully implantable platform at the system level (implant + leads/transceivers)

Electrode & Channel Physics

  • Channel / electrode count (single implant, reported): ~420–421 electrodes (reported across outlets)
  • Scalability (company claim): multi-implant linking to reach higher channel counts (exact numbers vary by statement)
  • Electrode material: platinum–iridium electrodes reported in secondary sources
  • Site area (µm²): not disclosed
  • Impedance @ 1 kHz: not disclosed
  • Noise floor / SNR: not disclosed
  • Recording modality: spikes / single-neuron activity emphasized in public reporting
  • Stimulation capability: not described as primary in public first-in-human reporting (treat as recording-focused)
  • Charge injection limit / safe stim range: N/A

Tissue Interface & Bioresponse

  • Target tissue: cortical gray matter
  • BBB disruption: moderate–high (penetrating)
  • Vascular disruption risk: moderate (implant-dependent)
  • Micromotion sensitivity: potentially significant (general issue for intracortical arrays)
  • Gliosis / encapsulation: expected over chronic timescales for penetrating interfaces; chronic data not yet publicly characterized for Connexus
  • Neuron loss (if reported): not disclosed
  • Foreign-body response mitigation: not disclosed
  • Typical failure mode: chronic stability/yield limitations typical of intracortical interfaces; hardware/connector/lead failures; infection risk if any percutaneous elements are used (goal is full implantation)

System Architecture

  • Onboard electronics: not fully specified in first-in-human reporting
  • Data path (acute test): intraoperative recording with external equipment implied
  • Telemetry (chronic concept): described as a fully implantable system with implanted transceiver(s) and external link in secondary sources
  • Sampling rate: not disclosed
  • Power: not disclosed
  • Thermal management: not disclosed
  • Hermeticity: not disclosed
  • MRI compatibility: unknown/conditional
  • Surgical complexity: intracortical placement during open cranial surgery
  • Implant/removal time (reported): removed intact in <20 minutes (press release)

Performance Envelope

  • Typical yield (acute): not disclosed quantitatively
  • Typical yield (chronic): not yet publicly reported (as of this entry)
  • Stability over time: not yet publicly reported (as of this entry)
  • Longevity (median / max): not yet publicly reported (as of this entry)
  • Revision / explant: acute explant demonstrated; chronic revision experience not yet publicly detailed
  • Adverse events (high-level): acute procedure reported as successful; chronic safety outcomes not yet publicly detailed
  • Notable demos / tasks: acute human neural recording during epilepsy surgery

Clinical / Preclinical Evidence

  • First-in-human context: temporary implant during epilepsy surgery; device implanted, recorded, and removed intact within minutes
  • N implanted subjects: at least one reported for acute test
  • Follow-up duration: acute only (minutes)
  • Indications (intended, chronic program): severe motor impairment / speech restoration (reported)
  • Trial registry links: not pinned here (ID not confirmed in-session)
  • Primary outcomes (acute): surgical feasibility, ability to record human neural signals, intact explant
  • Key limitations of evidence: acute-only public evidence; chronic performance and long-term safety require peer-reviewed and/or registry documentation

Engineering Verdict

Strengths:

  • high electrode count per implant (reported ~420–421)
  • demonstrated rapid implant + intact explant intraoperatively
  • clear translational pathway toward chronic, fully implantable communication systems

Limitations / failure modes:

  • acute-only human data disclosed so far; chronic stability/yield not yet published
  • detailed technical specs (geometry, impedances, yields) not yet available in a single primary technical paper

References