Robotic Rehabilitation Improves Sensorimotor Functions in Lower Limb Impairments of Individuals with Multiple Sclerosis

Background: People with multiple sclerosis (MS) often have symptoms of muscle weakness, spasticity, fatigue, imbalance, and sensory loss in the lower limbs, which may result in locomotion impairments.  The motor symptoms in MS are frequently accompanied by sensory deficits such as proprioceptive dysfunction resulting in a decreased motor control of the ankle and an increased risk of falls during ambulation. Therefore, rehabilitation strategies focusing on ankle recovery may improve mobility and safety of ambulation in patients with MS.   

Objectives: To investigate effectiveness of a robot-assisted ankle rehabilitation training protocol in reducing motor and sensory impairments and improving balance and gait functions.  

Methods: Six patients with MS participated in combined ankle passive stretching and active movement training using an ankle rehabilitation robot.  Participants received robotic training 3 sessions per week over 6 weeks for a total of 18 sessions.  Clinical and biomechanical evaluations were done at three time points: pre-treatment, post-treatment and 6-week follow-up evaluations. Proprioceptive acuity was measured as the plantar or dorsi-flexion movement (at the speed of 0.5°/sec controlled by the ankle rehabilitation robot) at which the patient felt the movement. 

Results: After six weeks of robot-assisted ankle intervention, there were significant increases in active range of motion in dorsiflexion (from -10.93±12.83° (a negative angle means not being able to reach the 0° dorsiflexion) before training to 11.26±8.90° after training), passive range of motion in dorsiflexion (from 6.19±4.10° to 15.6±6.06°), dorsiflexor muscle strength (from -0.94±9.52 Nm (a negative torque value means lower than the passive torque at 0° dorsiflexion) pre-training to 12.08±11.67 Nm post-training), and plantar flexor muscle strength (from 24.27±8.89 Nm to 34.97±8.01 Nm), with p < 0.05 for all above cases. Clinical assessments including Modified Ashworth Scale, Fugl-Meyer Assessment Lower Extremity score (FMA-LE), Berg Balance Scale, 6-minute walk test, and Timed Up-and-Go test showed trends of improvements after the training.  At the 6-week follow-up, the FMA-LE score showed significant improvement compared to pre-training (p=0.031).  The proprioceptive threshold reduced from 3.82±2.52° to 2.49±0.50° in dorsiflexion and from 4.09±3.45° to 3.07±1.23° in plantar flexion but the changes were not significantly different.

Conclusions: Robot-assisted training may be an effective therapeutic intervention to rehabilitate patients with MS for improving sensorimotor functions of the lower limbs.  The robotic training provided an interactive and motivating therapy to engage patients with MS in rehabilitation with improvements in sensory-motor performance. Further studies in a larger cohort are warranted to establish the value of these training techniques and the durability of the beneficial effects.