Researchers have presented a noninvasive, tongue-controlled hybrid hand system for people with cervical spinal cord injury. Combining a soft exoskeleton, tendon-driven flexion of all five fingers, and tongue-triggered functional electrical stimulation (FES), it supported daily-living grasps in a study of five users. Details of this study were published in npj Robotics.
Study: A tongue-controlled FES-robotic exoskeleton restores hand function in users with cervical spinal cord injury. Image Credit: Rawpixel.com
The Search for Better Hand Control
Spinal cord injuries affect over 20 million people worldwide, and cervical injuries often severely impair hand function, impacting the self-sufficiency of those affected.
Recent work has enabled hand movement through robotic exoskeletons and FES, while tongue-based interfaces have offered reliable, unobtrusive command input and remained functional after high-level injury.
However, key gaps remain in achieving reliable, non-invasive control for severe paralysis, enabling safe, dexterous hand opening and closing without added complexity, and mitigating FES-induced fatigue.
No tongue-based interface had yet been used to trigger FES or enable multiple grasps via a hand exoskeleton. This study addresses this with a non-invasive tongue-controlled hybrid FES-robotic hand exoskeleton.
Designing and Testing a Hybrid System
Five users with tetraplegia were recruited through clinicians at a spinal cord injury center in Western Denmark. Eligible participants were aged 18 to 75, had reduced or absent right arm motor function
from spinal cord injury, and could not repeatedly grasp and drink from a bottle while seated. Each had at least some tongue function.
The tongue interface sat in a custom dental retainer fitted to each participant's palate. A small titanium sphere on a wire frame acted as the activation unit, moved by the tongue across a grid of 15 inductive sensors to select commands. Sensor data was transmitted wirelessly to a central unit, then to a computer, which estimated the activation unit's position and drove the exoskeleton.
The hand exoskeleton was an adapted commercial soft glove, modified from three fingers to five and controlled to flex each finger independently by tendon. Force sensors at the fingertips detected object contact.
A passive-opening version added elastic wires on the backs of the fingers for extension, allowing comparison between two configurations: one with stimulation-driven opening and one with passive opening.
Stimulation used two constant current stimulators delivering brief pulses to the forearm extensor muscles. It was limited to finger extension and capped at two seconds to prevent fatigue.
Experiments were conducted over two consecutive days, starting with training and baseline testing without the exoskeleton, followed by grasping and force tasks with both configurations, and ending with a custom nine-item questionnaire.
Improved Grasping, Minimal Reported Discomfort
Without the exoskeleton, participants had limited or no ability to grasp independently. Two could lift light objects using wrist extension as a compensatory strategy, but performance was inconsistent, and none could reliably perform all grasp types.
One participant showed no observable arm or finger movement and could not grasp any object, and no participant could perform the pointing gesture.
With the exoskeletons, all five participants completed the tasks with high success. Overall success was 96% with the stimulation-driven opening exoskeleton and 89% with the passive-opening version.
Average completion times were similar, around 13 seconds for both. No participant required increased stimulation amplitude across trials, and no finger hyperextension was observed; extension was sufficient to release objects even for users with long-term injuries.
Saving this for later? Download a free copy now!
Questionnaire ratings favored the stimulation-driven opening exoskeleton for donning and doffing, willingness to use, and task difficulty. All participants reported the lowest possible discomfort and pain, and all but one reported the lowest fatigue.
Both exoskeletons significantly increased pinch and grasp forces compared with no exoskeleton. The difference between the two configurations was not statistically significant, though grasp forces trended higher with stimulation-driven opening.
The Promise and Limitations of Hybrid Assistance
The design separates roles, with the exoskeleton producing finger flexion and stimulation producing finger extension. This differs from typical hybrid systems where both act on the same joints.
By limiting stimulation to brief extension phases rather than continuous use, the approach lowers the risk of fatigue and keeps the system compact. The glove weighs 125 g on the hand, less than most comparable systems, though the full system, including control units, weighs about 1.155 kg.
The authors acknowledge that completion times of about 12 to 13 seconds remain higher than typical daily activities, though they include components beyond device actuation and are expected to decrease with training. They also note that fatigue ratings reflect perceived experience rather than measured physiological fatigue, and that objective fatigue assessment was not performed.
Limitations include a small sample size and portability constraints due to external hardware such as stimulators. Setup requires careful electrode placement and calibration for each individual.
Assistance is still needed for donning and doffing, and fingertip-only sensors meant that grasp detection sometimes relied on tendon force. The authors therefore suggest future work on additional sensors, shorter setup times, and possible therapeutic effects of stimulation.
Toward Restoring Everyday Hand Independence
This small study suggests that a tongue-controlled hybrid system can help people with cervical spinal cord injury perform everyday grasps while wearing it, including participants with long-term paralysis.
By pairing a soft exoskeleton for finger flexion with tongue-triggered stimulation for extension, the design enabled daily-living grasps with high success and low fatigue, while users rated it favorably for comfort and ease of use.
Journal Reference
Kirtas, O., et al. (2026). A tongue-controlled FES-robotic exoskeleton restores hand function in users with cervical spinal cord injury. Npj Robotics, 4(1). DOI:10.1038/s44182-026-00117-9. https://www.nature.com/articles/s44182-026-00117-9.
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.