Combining neural stimulation and a powered exoskeletal knee to enhance walking after stroke

Abstract

Introduction: In this study, we developed a unilateral, hybrid exoskeleton designed to augment post-stroke mobility through integration of noninvasive neural stimulation and a motorized knee orthosis. We evaluated proof of concept for this hybrid exoskeleton in a stroke survivor by assessing whether: (1) the motor and transmission possess sufficient backdrivability to facilitate effective muscle-generated movements, (2) the motor can independently drive movement, and (3) the combined application of stimulation and motor generate kinematic improvements superior to either modality used in isolation. Methods: The device was evaluated on an ambulatory stroke survivor under four overground walking conditions: (1) without the device, (2) stimulation only while wearing the exoskeleton, (3) motor assistance only, and (4) hybrid stimulated and motorized assistance. A motion capture system was used to obtain lower limb joint angles to assess the differences in each walking condition. Results: Changes towards neurotypical hip, knee and ankle angles were found between the gait with no device compared to the assistive conditions. Hip-knee cyclograms suggest hybrid assistance most resembled neurotypical gait. Conclusions: Individually, neural stimulation and motor assistance resulted in improved gait kinematics and hybrid assistance provided greater improvement than each form of individual assistance.

Keywords: Hybrid exoskeleton; kinematics; musculoskeletal; neural stimulation; stroke rehabilitation.

Figure 1.

User wearing the hybrid exoskeleton.

Figure 2.

Section view of knee actuator. Blue arrows indicate modification identified in Table 1.

Figure 3.

(a) Thigh cuff, (b) shank cuff, and (c) shank upright with their adjustment directions indicated by red solid arrows for anterior-posterior (A/P), green dashed arrows for proximal-distal (P/D), and blue dotted arrows for medial-lateral (M/L) adjustments, (d) Bottom view of the footplate with a red square representing its flexible forefoot.

Figure 4.

Sensor and electronics flowchart.

Figure 5.

(Upper) The four gait phases that comprise the finite state machine. (Lower) Targeted muscles and motor assistance applied for each phase.

Figure 6.

Hip, knee, and ankle joint angles for (a) no device (ND), (b) stim only (S), (c) motor only (M), and (d) hybrid (H) walking conditions. Solid lines represent the mean angle while gray shading represents ±1 standard deviation. Solid vertical lines represent mean location of foot off with the dashed vertical lines representing ±1 standard deviation.

Figure 7.

Hip–knee cyclograms, moving clockwise for (a) no device (ND), (b) stim only (S), (c) motor only (M), and (d) hybrid (H) walking conditions. The solid lines represent stance phase, beginning at the black diamond dashed lines represent swing phase, beginning at the black star. The black circle represents the point where the foot crosses immediately under the pelvis.

Figure 8.

95% confidence intervals for the joint angle means at initial contact (IC), foot off, and peak values during swing for the hip, knee, and ankle for each condition.