Neural Noise- A peripheral neuroprosthesis can support gastrointestinal motility and metabolic neuromodulation (Nature).1
- The work is relevant to neuroprosthetics and device-based neuromodulation but not CNS BCI.1 1
Weekly enrichment (2026-07-20)
- The primary study is Srinivasan, Antonini, Alshareef, Sahasrabudhe et al. (MIT and Tufts groups, including Robert Langer and Polina Anikeeva), published in Nature Communications in 2025 as a closed-loop implant that interfaces directly with the enteric nervous system.2
- The neuroprosthesis is a fiber-drawn device: seven stainless-steel stimulation electrodes surround a central microfluidic drug-delivery channel inside a biocompatible polycarbonate cladding, at a 1.25 ± 0.10 mm diameter chosen to pass through the 2.8–3.2 mm working channels of standard endoscopes, with contacts spaced 1 cm apart to mirror myenteric nerve-root arborization.2
- For the oesophagus, optimized electrical stimulation of 40 Hz, 2 mA, 100 µs pulse width and 0.5 s trains produced complete luminal closure, and a frequency sweep at 2 mA showed 40 Hz gave sustained 89 ± 4% closure (n = 4 animals, 10 repetitions per condition).2
- In a paretic-stomach swine model, gastric stimulation at 20 Hz/2 mA roughly doubled the motility rate (measured as pyloric closure cycles proportional to peristalsis), a significant increase over baseline (p < 0.01, Student’s t-test, n = 4 animals).2
- Closed-loop food sensing was implemented by differential amplification of electrode impedance at 0.5, 1, and 2 kHz, detecting ingested boluses at forces greater than 0.2 N; after each stimulation train a 5 s refractory period mimicked natural deglutitive inhibition.2
- In a metabolic experiment, 30 minutes of oesophageal stimulation in fasted swine (n = 6 stimulated vs. n = 6 unstimulated controls) recapitulated a satiated hormone profile with low glucagon and ghrelin plus an insulin response consistent with food intake, whereas passive mechanical stretch alone did not elicit the response.23
- Durability and safety: accelerated aging applied 144 × 10⁶ biphasic pulses (approximately 8 years of anticipated use) with electrodes remaining below the −1.25 V water-reduction potential, and biocompatibility testing in rats (n = 14) and swine showed only a thin fibrous capsule and no substantial foreign-body reaction.2
- Positioning: the device targets the peripheral enteric nervous system rather than a CNS brain–computer interface, and all experiments used an anaesthesia-induced dysmotility model in swine — no human clinical data are reported yet.2
- Context: the authors note that approved GI electroceutical devices (EnteroMedics Maestro VBLOC for obesity, Enterra gastric neurostimulator for gastroparesis, IntraPace abiliti, Medtronic InterStim) have shown inconsistent effects on gastric emptying, and reviews of bioelectric neuromodulation similarly report variable clinical efficacy, motivating the closed-loop multichannel approach.3