Fig. 2.

Tracking task performance during single-DOF IN-type tests. Exoskeleton joint angular position, electromyogram (EMG) data and model-based estimates of joint moments are reported during tasks with one degree of freedom (DOF). Data are reported as averages across all tracking trials with standard deviations (shaded area). They are reported for the low-gain (LG) and high-gain (HG) exoskeleton assistance levels and as a function of percent cycle, i.e., where 0 and 100%, respectively, represent the beginning and the end of the tracking trajectory (Target). Target trajectories are personalized to each patient (Table 3) as detailed in the Methods section. The results are relative to tests inside the exoskeleton, i.e., IN-type tests. Data are reported for two representative healthy subjects (Healthy 1–2), two stroke patients (Stroke 1–2) and one incomplete spinal cord injury (SCI) patient. The results are reported both for the individual control of the exoskeleton ankle plantar-dorsiflexion DOF and for that of the exoskeleton knee flexion-extension DOF. EMGs are relative to muscles, including the biceps femoris (BF), rectus femoris (RF), semimembranosus (S), vastus lateralis (VL) and vastus medialis (VM), soleus (So), gastrocnemius medialis (Ga) and tibialis anterior (TA), as shown in Table 1