SmartVNS Clinical Feasibility Study

26 September 2025

Overview of the SmartVNS system used in this study. The device includes three parts: (A) a lightweight horseshoe-shaped stimulator worn around the neck, with simple plus and minus buttons to adjust intensity and lights to show status; (B) a small, optimized ear earpiece to deliver stimulation; and (C) a wrist-worn motion sensor that detects movements and automatically triggers stimulation at the right time.

We are excited to share that our team has just submitted a manuscript presenting SmartVNS — a fully wearable system for closed-loop non-invasive neuromodulation in neurorehabilitation.

Read the preprint here: Closed-Loop Movement-Paired Transcutaneous Auricular Vagus Nerve Stimulation for Upper-Limb Rehabilitation: a Feasibility Study

SmartVNS combines:

A wrist-worn inertial sensor for real-time detection of functional movements
A custom stimulator and optimized auricular electrode for transcutaneous vagus nerve stimulation (taVNS)
Automated movement-triggered stimulation, removing the need for therapist manual control

In a 4-week clinical feasibility study (20 sessions, 9 participants with stroke or spinal cord injury), SmartVNS demonstrated:

Consistent movement-paired stimulation across patients with diverse impairment levels
Comparable precision to manual therapist-triggered stimulation, with nearly double the sensitivity
High usability ratings (UMUX ~85%) from both patients and therapists
Simple and fast setup: patients self-applied the system in under 2 minutes, even with significant upper-limb hemiparesis
Preliminary evidence of modest functional improvements in upper-limb outcomes, supporting the need for further controlled studies

These results suggest that SmartVNS could become a scalable and patient-friendly adjunct to conventional neurorehabilitation.

Huge thanks to our collaborators, our clinical partner cereneo, and participants who made this possible.

Authors: Clément Lhoste, Max Quast, Andrea Ronco, Abigail Vogel, Chris Easthope Awai, Meret Branscheidt, Olivier Lambercy, Paulius Viskaitis, Dane Donegan

A stroke survivor is shown setting up the SmartVNS device on their own using only their stronger hand. The images show the quick setup process: starting the setup, placing the stimulator around the neck, positioning the small electrode on the ear, and adjusting the stimulation to a comfortable level. A timer in minutes and seconds highlights how fast the setup can be completed (about 1 minute).

Stimulation logic provided similar range of stimulation dosage among patients with different abilities (baseline ARAT scores) and performing different type of exercises. Each boxplot represents a single patient (n=9), with individual data points corresponding to the average stimulation rate recorded during a single therapy session.

Comparison of two ways to trigger stimulation during therapy: a therapist pressing a button versus the SmartVNS wrist sensor automatically detecting movements. The first panel shows an example with movement data, stimulations delivered by the device, stimulations triggered by the therapist, and the true movements from video recordings. The second and third panels show how well each method worked: precision (how often stimulations matched real movements) and sensitivity (how many real movements were successfully paired with stimulation). Each point represents one therapy session, with lines connecting therapist and device performance. The automated SmartVNS approach showed higher sensitivity while maintaining similar precision.

❮ ❯