Rehabilitation often depends heavily on visual demonstrations, screen-based instructions, and verbal reminders. While these methods can help patients complete a task, they may also draw attention away from body awareness and make long-term learning overly dependent on external guidance. Brain–computer interface (BCI) systems have also been explored for rehabilitation, but asking them to provide continuous, fine motor control remains difficult in everyday wearable settings because brain signals are often noisy and can vary across users and across days.
In a new article, the authors propose a more restrained and practical design concept. Instead of using electroencephalography (EEG)-based BCI to continuously command movement, the system uses brain activity as an intention gate—a signal that detects when a patient is trying to move, and then decides whether haptic guidance should be delivered and at what intensity. This design shifts the role of BCI from direct control to timed support. At the same time, wearable haptic modules based on McKibben artificial muscles provide simple directional cues, such as left/right, forward/backward, or up/down, helping users adjust their movement in an intuitive “coach-like” way.
The proposed framework emphasizes safety, simplicity, and skill transfer. Because the brain signal only gates cue delivery rather than directly driving body movement, occasional decoding uncertainty is less likely to cause harmful motion commands. The design also favors a small and stable set of haptic meanings, so that each cue remains easy to interpret across training sessions. Over time, cue intensity can be gradually reduced, encouraging patients to rely more on their own perception and motor planning rather than on the device.
The concept is especially relevant for neurological rehabilitation, such as stroke recovery, where patients may still have detectable movement intention but need help improving movement quality and reducing dependence on external prompts. The authors suggest that this design philosophy could support future closed-loop rehabilitation systems by focusing less on maximizing robotic assistance and more on delivering the minimum necessary, intention-timed guidance needed for effective learning. The work titled “A rehabilitation design concept based on brain–computer interface and McKibben artificial muscle” was published in Healthcare and Rehabilitation (published on March 18, 2026).

DOI: 10.1016/j.hcr.2026.100066