Wristful thinking: exploring the effects of robotic rehabilitation and cross-education for persons with multiple sclerosis

Abstract

The goal of this work was to develop an adaptive rehabilitation technique using a haptic wrist robot that would induce cross-education to an untrained limb. Fourteen individuals with Multiple Sclerosis (MS) and eight non-affected adults completed an eight-week intervention. MS participants were placed into two groups, training their more affected limb (direct training), and training their less affected limb (indirect training). The purpose of the intervention was to improve wrist and grip strength (measured via maximal grip and isometric wrist strength) and motor control (measured via robotic assessments). Participants trained with the robotic device three times per week for eight consecutive weeks. Training consisted of eccentric contractions as the participant resisted a force elicited from the robotic device as it moved in flexion, extension, and radial-ulnar deviation. The MS group reported significant increases in wrist strength. The indirect training group significantly improved in flexion, extension, radial and ulnar deviation in the trained limb, and flexion and radial deviation in the untrained limb. The direct training group showed improvements in extension and ulnar deviation in both limbs. The control group improved in radial and ulnar deviation, with radial deviation improving in the untrained limb. Grip force remained unchanged for all groups. MS participants significantly decreased tracking and figural error post-intervention suggesting evidence that motor control adaptations occurred following an adaptive and resistive robotic intervention of the upper limb. Results of this work provide evidence that eight-week robotic rehabilitation can elicit cross-education effects to the untrained limb.

Keywords: Cross-education; Multiple sclerosis; Neurorehabilitation; Robotic rehabilitation; Upper extremity.

Fig. 1

Timeline of intervention. Abbreviations: IT: indirect training, DT: direct training, CON: control, DB: double baseline

Fig. 2

(A) Participant positioned in the robotic device prepared for a training session. (B) Path of the moving target icon in eccentric flexion. Crosses indicate the four directions of movement (flexion, extension, radial and ulnar deviation)

Fig. 3

Strength (N) PRE and POST for each group (A) IT, (B) DT, (C) CON for the trained and untrained limbs and (D) DB, for the dominant and non-dominant limbs. Data are presented as mean scores (solid bars), standard deviation (black error bars) and individual data points (circles) with the corresponding POST value. An asterisk denotes a significant change from PRE to POST. Abbreviations: F: flexion, E: extension, RD: radial deviation, UD: ulnar deviation, IT: indirect training, DT: direct training, CON: control, DB: double baseline

Fig. 4

Trajectories displayed for the tracking task PRE (blue) and POST (grey) for each group (A) IT, (B) DT, (C) CON for the trained (left) and untrained limbs (right), and (D) DB, for the dominant and non-dominant limbs. Data represents the target trace (black) and the mean performance per lap across each group. X axis represents flexion/extension and Y axis represents radial and ulnar deviation. X and Y axis denotes angular displacement in radians at the wrist. Asterisk depicts significance for tracking error and the plus symbol depicts figural error. Abbreviations: IT: indirect training, DT: direct training, CON: control, DB: double baseline, TE: tracking error, FE: figural error