Technology-Assisted Rehabilitation: Empowering Active Participation in Training and Early Recovery of Ambulatory Function
As populations around the world rapidly age, mobility health is faced with unprecedented challenges. With long-term funding from the National Science and Technology Council (NSTC), an interdisciplinary team led by Distinguished Associate Professor Lian-Wang Lee of National Chung Hsing University (NCHU), Professor Li-Wei Ko of National Yang Ming Chiao Tung University (NYCU), and Professor Chia-Hsin Chen of Kaohsiung Medical University (KMU) has collaborated for ten years to develop the world’s first Brain-Reading AI-Driven Walking Rehabilitation Robot (HopeStride). This intelligent rehabilitation robot—capable of “reading the brain”—now allows stroke survivors and patients with degenerative conditions to execute “brain-driven walking.” It enables active participation, accelerates recovery, reduces therapist workload, and enhances clinical efficiency for rehabilitation physicians.
The system integrates automatic vehicle assist, dynamic body-weight support, an exoskeleton, and a noninvasive wireless brain–computer interface (BCI). It analyzes brain signals to interpret motor intentions, then drives the exoskeleton to deliver safe and natural gait training. The same device supports standing, balance training, gait guidance, natural walking, and brain-driven exoskeleton locomotion. This represents a breakthrough from limitations of passive training and can be flexibly utilized across hospitals and rehabilitation centers.
This Active Mode rehabilitation approach helps patients rebuild neural pathways in a more natural gait state—a major breakthrough that is difficult to achieve with traditional equipment. Clinical trials at KMU Chung-Ho Memorial Hospital have demonstrated significant gains in patients’ gait control, lower-limb strength, and cardiopulmonary endurance. The findings highlight how the system could be deployed more widely to help more people reclaim their mobility.
Passive Rehabilitation Fails to Trigger Adequate Brain Activation, while Brain-Driven Walking Introduces a New Training Paradigm
For many patients, the most frustrating part of rehabilitation is that walking again feels like a distant dream. Traditional rehabilitation devices, while effective at reducing fall risk by providing mechanical support for standing and walking, are largely passively driven. This passive mode often results in rigid gait patterns, restricted body swing, and an inability to replicate natural walking rhythms. Most importantly, these systems rarely succeed in actively engaging brain activity or brain–limb coordinated control. Furthermore, these devices often provide only single-function support, lack real-time feedback, and offer no automation.
As a result, therapists cannot easily adjust training parameters in response to a patient’s immediate performance, limiting both the efficiency and safety of the rehabilitation process. This passive approach—where the body moves but the brain does not truly learn—makes meaningful functional reconstruction difficult.
Therefore, clinicians and engineers have been working toward a new generation of smart rehabilitation systems that integrate motor training with neural remodeling. The goal is to shift from passive training to active, brain-driven gait regulation—leveraging neural plasticity to restore the brain–body connection. This marks an important breakthrough for clinical rehabilitation.
A Decade of Refinement: Building a "Brain-Reading" Rehab Robot
Supported by the NSTC since 2016, the cross-institutional research team has developed a Brain-Reading AI-Driven Walking Rehabilitation Robot (HopeStride) built upon an integrated concept of neural plasticity, safety control, artificial intelligence, and BCI technologies. The system dynamically adjusts its training mode to match different stages of rehabilitation:
■Early phase: Assists with standing and balance while activating lower-limb movement.
■Middle phase: Uses vision-assisted gait guidance to establish natural walking patterns and proper weight shifting.
■Late phase: Actively drives the exoskeleton via electroencephalogram (EEG) signals, achieving "intention-controlled walking".
The system stimulates both the central and peripheral nervous systems, forming a central–peripheral–central closed-loop training cycle. This architecture meets clinical demands for a rehabilitation training program that is active, personalized, modular, and precise.
Back to Daily Life Is No Longer Just a Dream: Clinical Trials Support the Benefits of Brain-Driven Walking
At the Rehabilitation Department of KMU Chung-Ho Memorial Hospital, the research team conducted a study comparing two groups across 20 treatment sessions:
■ Control group: received passive exoskeleton training
■ Experimental group: received BCI-based intention-driven exoskeleton training
Assessments of balance, limb function, and cardiopulmonary endurance showed that the control group improved primarily in lower-limb function. In contrast, the experimental group demonstrated concurrent improvements in both lower-limb function and cardiopulmonary endurance. These findings indicate that brain-driven rehabilitation enhances both motor performance and neural engagement, bringing about dual benefits.
The NSTC, through its Research Project of the Control Engineering Program and the International Young Distinguished Scholars Program, continues to foster interdisciplinary research and accelerate the translation of smart rehabilitation technologies from the lab to the bedside. The team plans to complete more than 50 clinical validation cases by the end of 2026. This includes integrating the proprietary wearable wireless BCI developed by Professor Li-Wei Ko of NYCU, which has been certified for medical-device safety and biocompatibility testing. The team will also continue to refine the robot’s design and clinical applications.
This groundbreaking achievement not only addresses the pressing rehabilitation challenges of an aging society but also reflects the core mission of using technology to bridge manpower gaps. As rehabilitation robotics and BCI technologies continue to advance, intention-based walking is poised to become a new benchmark in smart medicine—turning the hope of returning to daily life into a tangible and achievable future.
References:
1. Lian-Wang Lee*, I-Hsum Li, Ting-Wei Liang, “A Proof of Concept Study for the Design, Manufacturing, and Control of a Mobile Overground Gait-Training System,” International Journal of Fuzzy Systems, vol. 23, pp. 2396-2416, 2021.
2. Chun-Ren Phang, Kai Hsiang Su, Yuan Yang Cheng, Chia Hsin Chen, Li Wei Ko*, “Time synchronization between parietal–frontocentral connectivity with MRCP and gait in post stroke bipedal tasks,” Journal of NeuroEngineering and Rehabilitation, vol. 21, pp. 1-16, 2024.
3. Lian-Wang Lee, I-Hsum Li*, “Safety-enhanced control for a MuscleDrive waist-assistive exoskeleton,” Control Engineering Practice, vol. 156, 106182, 2025.
Media Contact：
Ching-Chun Tu, Program Manager
Department of Engineering and Technologies, National Science and Technology Council
Phone: +886-2-27377527
Email: cctu@nstc.gov.tw
