Devices

A fully implanted, wireless epidural ECoG system developed by CEA‑Clinatec for human motor BCI research, emphasizing clinical robustness and long-term stability over single-unit precision.

Device — Cortical surface

WIMAGINE (CEA‑Clinatec) epidural wireless ECoG implant

Official site →

BCI · ECoG · epidural · wireless · CEA · Clinatec · WIMAGINE · motor decoding · cortex · recording

WIMAGINE (CEA‑Clinatec) epidural wireless ECoG implant

One-line verdict: A fully implanted, wireless epidural ECoG system that trades spatial precision and single-unit access for a safer surgical interface class and clinically robust long‑term recordings for human motor BCIs.

Quick tags: Recording · Closed-loop capable (system-dependent) · Species: Human · First implanted: ~2016

Overview

What it is: WIMAGINE is a fully implanted, wireless epidural ECoG brain–computer interface developed by CEA‑Clinatec for human motor restoration research. It records population-level cortical signals over motor cortex without penetrating brain tissue and transmits data wirelessly to external decoding systems.

Why it matters: WIMAGINE is a concrete demonstration that stable, long-term motor decoding is possible in humans using a non‑penetrating, fully implanted neural interface class — enabling high-profile demonstrations (e.g., exoskeleton control) while reducing biological risk relative to penetrating arrays.

Most comparable devices: subdural/epidural ECoG systems (signal class), other fully implanted wireless cortical surface interfaces.

Spec Card Grid

Identity

  • Device name: WIMAGINE
  • Canonical ID: BTSD-IMBCI-0004
  • Inventor / key authors: Alim‑Louis Benabid et al.
  • Org / manufacturer: CEA‑Clinatec (France)
  • First demonstrated (year): ~2016 (reported)
  • First implanted (year): ~2016 (reported)
  • Species: human
  • Regulatory / trial status: human research trial
  • Primary use: recording
  • Primary target: motor cortex (epidural over M1)

Geometry & Architecture

  • Interface type: epidural ECoG
  • Penetrating?: no
  • Form factor: flexible electrode array
  • Array layout: multi-contact ECoG grid
  • Footprint (mm): cm-scale cortical coverage (reported)
  • Insertion depth (mm): epidural (no cortical penetration)
  • Shank / lead dimensions: N/A
  • Site spacing (µm): mm-scale
  • Tip geometry: flat contacts
  • Insertion method: craniotomy with epidural placement
  • Anchoring method: skull fixation
  • Packaging location: fully implanted cranial module with wireless telemetry

Electrode & Channel Physics

  • Channel count: ~64 (reported; configurations vary)
  • Active sites used (vs total): majority active (reported)
  • Electrode material: platinum-based contacts (reported)
  • Site area (µm²): large surface contacts (ECoG-scale)
  • Impedance @ 1 kHz: low (typical for ECoG; exact values vary)
  • Noise floor / SNR: moderate; population-level signals
  • Recording modality: ECoG (LFP-dominant)
  • Stimulation capability: no (recording-focused)
  • Charge injection limit / safe stim range: N/A

Tissue Interface & Bioresponse

  • Target tissue: dura-adjacent cortical surface
  • BBB disruption: low
  • Vascular disruption risk: low
  • Micromotion sensitivity: low
  • Gliosis / encapsulation: generally lower than penetrating arrays; long-term encapsulation can still affect coupling
  • Neuron loss (if reported): none reported (in summary sources)
  • Foreign-body response mitigation: non-penetrating geometry
  • Typical failure mode: coupling changes (encapsulation), hardware aging, telemetry/power issues

System Architecture

  • Onboard electronics: amplification + digitization
  • Data path: fully implanted wireless telemetry
  • Telemetry bandwidth: sufficient for ECoG decoding (exact figures vary)
  • Sampling rate: ECoG-appropriate (often hundreds of Hz to kHz in practice; exact varies)
  • Power: implanted power module (inductive recharge reported)
  • Thermal management: within safe cranial limits (reported)
  • Hermeticity: medical-grade sealed implant
  • MRI compatibility: unknown/conditional (device- and protocol-dependent)
  • Surgical complexity: moderate craniotomy, low cortical risk (relative to penetrating arrays)

Performance Envelope

  • Typical yield (acute): high
  • Typical yield (chronic): stable multi-month recordings (reported)
  • Stability over time: good
  • Longevity (median / max): multi-year implants reported in this program
  • Revision / explant: feasible
  • Adverse events (high-level): none major reported in summary sources
  • Notable demos / tasks: robotic exoskeleton control; motor intention decoding in tetraplegia

Clinical / Preclinical Evidence

  • N implanted subjects: small human cohort
  • Follow-up duration: months to years
  • Indications: motor paralysis / tetraplegia
  • Trial registry links: NCT02550522
  • Primary outcomes: feasibility of wireless epidural motor BCI
  • Key limitations of evidence: limited spatial resolution; small sample sizes; device specs not uniformly reported

Engineering Verdict

Strengths:

  • strong safety profile relative to penetrating arrays
  • fully implanted, wireless human system
  • clinically realistic surgical burden

Limitations / failure modes:

  • limited spatial and single-unit resolution
  • decoding ceiling for surface/population signals

Scaling constraints:

  • information density per unit area
  • telemetry/power budget
  • long-term coupling changes (encapsulation)

What newer designs try to fix:

  • higher channel counts
  • improved spatial specificity without penetrating tissue

References