title: Graphene transistors and stroke-tissue signatures description: What the September 2026 Brain paper actually measured with graphene microtransistor arrays in mouse stroke models, and what the October press added. date: 2026-10-08 tags:

  • engineering
  • medical
  • research

Evidence: Well-supported for the mouse electrophysiology and imaging results in Flaherty et al., Brain (2026). Not a clinical device and not treatment advice.

Related notes: biomechanical · glossary · 2026

If you only remember five facts

  1. The sensor is a graphene solution-gated field-effect transistor (gSGFET) array: CVD graphene on a flexible probe, electrolyte as the gate. Not graphene oxide, not a passive metal electrode.
  2. Flaherty et al., Brain, online 9 September 2026, DOI 10.1093/brain/awag305. Manchester press and ICN2 posts landed 7–8 October. The DOI is older than the headline.
  3. In mouse focal ischaemia, cortical spreading depolarization (CSD) waveform duration and shape scaled with local perfusion, mapped against optical blood flow and oxygenation.
  4. A characteristic waveform marked profoundly hypoperfused tissue and predicted whether vessels dilated, went biphasic, or constricted.
  5. Low-dose ketamine shortened CSDs in that study and reduced injury size in the press account of the paper. Study result only. Not a dosing guide.

What was measured

High-density graphene micro-transistor arrays sat on cortex while optical imaging read blood flow and oxygenation. Two complementary focal-ischaemia models. Authors include University of Manchester (Flaherty, Wykes, Allan), IMB-CNM CSIC (Guimerà-Brunet, Illa), ICN2 (Garrido and colleagues), and Multi Channel Systems (Meents, Jeschke, Mohrlok). Kostas Kostarelos is on the author list.

The abstract’s claim: CSD signatures resolve a continuum from relatively preserved cortex to markedly ischaemic regions, and electrical features predict the haemodynamic response. Oxygenation imaging showed a focal pre-onset drop at initiation sites. Prolonged CSDs left a sustained oxygen burden.

Press language (“real-time map for doctors”) is prospective. The data are mice.

Why graphene transistors

Spreading depolarizations are slow, near-DC shifts. Passive electrodes and AC-coupled amplifiers lose that band. A gSGFET converts local potential to channel current and can be DC-coupled. The Barcelona stack (CVD graphene on polyimide, solution gate) is the same family as Masvidal-Codina et al., Nature Materials 2019, and later multiplexed arrays (Schaefer et al., 2D Materials 2020, DOI 10.1088/2053-1583/ab7976).

This paper is not the INBRAIN first-in-human cortical study NCT06368310 (tumor-surgery safety). Related groups, different protocol. It is also not the 7 October 2026 MINIGRAPH robot-implantation press (large animal and bench; company statement).

Limits

  • Mouse preparation with an optical window. No human stroke recording in this paper.
  • Ketamine finding is mechanistic inside the study. Not medical advice.
  • Transfer quality, bias stability, and chronic biofouling are separate engineering problems.
  • Do not merge with graphene-oxide fringe claims. Wrong material.

Sources