Feasibility, Safety and Efficacy of ZEPU AI9 Lower Limb Exoskeleton Rehabilitation Robot in Patients With Lower Limb Motor Dysfunction

Lower-limb motor dysfunction following stroke, incomplete spinal cord injury, and orthopedic surgery is a major contributor to disability in low- and middle-income countries, where access to intensive, repetitive gait training is often limited by shortages of rehabilitation professionals and structured multidisciplinary programs. Robotic-assisted gait training has emerged as a way to deliver consistent, task-specific, and intensity-controlled practice that is difficult to achieve with therapist-delivered conventional therapy alone, by enabling programmable gait cycles with controlled step length, cadence, symmetry, and body-weight support.

The ZEPU-AI9 is a dual-leg, passive-training lower-limb exoskeleton (Shandong Zepu Medical Technology Co., Ltd.) that generates repeated physiological gait cycles through motorized hip and knee joints while maintaining ankle alignment within a controlled range. It supports standing, sitting, stepping-in-place, and over-ground walking modes via a touchscreen interface and handheld steering joystick, with adjustable thigh/calf length, height, waist width, joint movement angles, step length, and training duration to accommodate individual anthropometrics. Built-in safety features include an emergency stop, automated spasm protection, mechanical joint limits, overload detection, low-battery alarms, and anti-tilt protection, supplemented by an overhead fall-arrest harness during supervised sessions. Powered lower-extremity exoskeletons are classified as Class II medical devices by the U.S. FDA (product code PHL), reflecting both their therapeutic potential and the need for supervised clinical use and standardized protocols. No clinical data on the ZEPU-AI9 currently exist from Bangladesh, and its performance within a Bangladeshi rehabilitation setting has not been evaluated.

This single-site pilot randomized controlled trial will be conducted at the Robotic Rehabilitation Centre, Department of Physical Medicine and Rehabilitation, Bangladesh Medical University. Following screening and informed consent, eligible participants will be randomized to receive either ZEPU-AI9-assisted gait training in addition to standard-of-care rehabilitation, or standard-of-care rehabilitation alone, over a 12-week intervention period. Exoskeleton sessions will be conducted three times weekly (45 minutes each, comprising warm-up, device donning, structured gait training, and cool-down), supervised by physiatrists and physiotherapists trained in robotic rehabilitation, with training parameters progressively adjusted based on participant performance and tolerance. Pre- and post-session safety checks (vital signs, skin integrity, pain/fatigue via VAS) will be performed at every session, with device-related adverse events recorded in real time.

Standardized assessments (6-Minute Walk Test, Manual/MRC Muscle Strength Testing, Functional Ambulation Category, Berg Balance Scale, Timed Up and Go test, Modified Ashworth Scale, and a validated Bengali quality-of-life measure) will be administered at baseline and at two-week intervals through week 12, allowing paired comparison of pre- and post-intervention values. Feasibility will additionally be evaluated through recruitment and screening metrics, session adherence and dropout rates, therapist training time, device set-up and donning/doffing time, and a Post-Study System Usability Questionnaire administered at study completion.

Data will be analyzed on an intention-to-treat basis (all randomized participants analyzed per original group assignment), with a secondary per-protocol analysis restricted to participants completing the intervention as specified. Continuous variables will be summarized as mean ± SD (or median/IQR where non-normally distributed) and compared using paired or unpaired t-tests, or Wilcoxon signed-rank tests as appropriate, with statistical significance set at p<0.05. Findings from this pilot study are intended to generate feasibility indicators and preliminary effect-size estimates to inform the design of a future, adequately powered randomized controlled trial and to guide national implementation strategies for robotic rehabilitation in Bangladesh.