NeuroString stretchable neurochemical sensor

Li and colleagues’ 2022 device senses neurotransmitters using an interconnected graphene/metal-oxide nanoparticle network embedded in an elastomer. Stanford’s institutional account identifies the work as originating in Zhenan Bao’s group. The paper lists Stanford and Michigan State affiliations among its authors.

NeuroString measures chemical dynamics. It is not an intracortical spike-recording array, a stimulation implant or an established human BCI.

Fabrication and sensing interface

Laser patterning carbonizes a metalloporphyrin-containing polyimide precursor to form graphene with iron-oxide or nickel-oxide nanoparticles. The network is transferred into SEBS elastomer, encapsulated and cut into strings. Exposed cross-sectional tips provide the sensing surface. Extended Data Figure 1 describes an additional elastomer coating, oxygen-plasma exposure and Nafion coating to reduce ascorbic-acid interference.

The brain version uses a temporary pullulan coating for insertion. Dissolution releases the soft strings. The gut version has a distinct layout and preparation for placement within a moving lumen.

Published layout distinctions

The primary figures show a three-channel brain sensor. Extended Data Figure 1’s brain layout labels a 3-5 mm intracerebral segment and ranges of 100-150 µm and 50-150 µm for its lateral dimensions. Its fabrication caption separately describes cutting individual strings to 90 µm width. These are kept as different drawing and process specifications rather than collapsed into one universal dimension.

The gut layout labels a 30-200 mm segment, approximately 200 µm and a 100-200 µm lateral range. Those dimensions are not a brain-implant specification. The full stack thickness and a complete coordinate map are not established by the sources used here, so no 3D model is supplied.

Characterization versus in vivo performance

Figure 1 tests the material under deformation and up to 5,000 cycles at 0-50% strain. That material test does not establish 5,000 successful in vivo sensing sessions. Figure 2 compares electrochemical selectivity and calibration in solution, including dopamine, serotonin, epinephrine and norepinephrine. Solution calibration and changing pH are relevant limits on interpreting estimated in vivo concentrations.

The linked brain and gut experiments include mouse reward learning, optogenetic and pharmacological perturbations, and gut serotonin measurements. Some results are labeled catecholamines rather than uniquely dopamine; the entry preserves that distinction.

Translation boundary

Stanford’s 2022 account says the implant was not ready for clinical use and remained wired to external readout. Potential uses in Parkinson’s disease, depression and intestinal disorders are research directions, not clinical outcomes or indications. The primary abstract describes chronic sensing, but this entry does not assign a specific chronic lifetime without its full duration evidence.

Primary sources

• 2022 primary paper and abstract.
• Figure 1: structure and deformation tests.
• Figure 2: electrochemical characterization.
• Extended Data Figure 1: fabrication and distinct brain/gut layouts.
• Stanford research account and clinical-readiness caveat.