Related Experiment Video
Updated: Mar 25, 2026
A Rehabilitation Program of Exoskeleton-assisted Body Weight-Supported Treadmill Training with Non-immersive Virtual Reality for Stroke Patients
Published on: May 16, 2025
Identifying candidates for torque-assisted exoskeleton for gait assistance after stroke: a pre-specified subgroup
Won H Chang1, Tae-Woo Kim2, Hyoung S Kim3
1Department of Physical and Rehabilitation Medicine, Center for Prevention and Rehabilitation, Heart Vascular and Stroke Institute, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea.
Background:
Wearable robots show promise for gait assistance in stroke patients, yet the clinical characteristics predicting a positive ambulatory response to exoskeletal assistance remain unclear.
Aim:
To identify appropriate candidates for torque-assisted exoskeletal wearable robots in stroke patients.
Design:
A subgroup analysis using data from an international, multicenter, randomised controlled trial.
Setting:
Inpatient.
Population:
Fifty-five early subacute stroke patients who completed four weeks of robot-assisted gait training (RAGT) with the wearable exoskeletal robot (ANGEL LEGS M20, Angel Robotics Co., Ltd.).
Methods:
Immediately after RAGT for the four weeks, ambulatory function with the exoskeleton on and off was evaluated using the 10-Meter Walk Test (10MWT), 6-Minute Walk Test (6MWT), and Physiological Cost Index (PCI). At the same time, additional assessments included the Functional Ambulatory Category (FAC), Fugl-Meyer Assessment-Lower Extremity, Motricity Index-Lower Limb, Trunk Control Test, and Berg Balance Score. Participants were classified as good-responder, no-responder, or negative-responder groups based on changes in walking performance with exoskeletal assistance. Univariate and multivariate ordinal logistic regression analyses identified factors associated with responsiveness.
Results:
In the good-responder group, 10MWT, 6MWT and PCI showed significant improvements in the robot-on state compared with the robot-off state, respectively (P<0.05). Good responders had significantly lower baseline ambulatory, balance, and lower limb motor function compared to negative-responders (P<0.05). Multivariate analysis identified lower FAC as the only independent predictor of positive response to exoskeletal assistance (P<0.05).
Conclusions:
Torque-assisted exoskeletal wearable robots may improve ambulatory function in stroke patients with low ambulatory function.
Clinical Rehabilitation Impact:
To achieve meaningful effects through exoskeleton robots, patient selection must be adjusted according to clinical needs.
