Effects of Visual and Verbal Feedback on Active Patient Participation During Robot-Assisted Gait Training

Robot-Assisted Gait Training (RAGT) is a widely adopted neurorehabilitation approach for individuals with Spinal Cord Injury (SCI). In this context, active patient participation is a key factor to improve the success of the therapy. Previous works with the robotic trainer Lokomat have shown that realtime visual feedback and verbal instructions from a physical therapist effectively enhance active participation. However, the optimal type of visual feedback and its potential synergy with verbal indications remain unclear. This study investigates how different visual feedback types influence active participation when combined with therapist verbal instructions. Nine participants with SCI underwent RAGT sessions with three types of visual feedback (chart, emoticon, game) and two levels of patienttherapist verbal interaction (high, low). Active participation was quantified from a biomechanical point of view through the human-robot interaction work in four joints (left/right hip and knee) during the stance and swing phases of the gait cycle. Results showed that feedback chart was the most effective in promoting positive, symmetric performance, as opposed to feedback game and emoticon which, though probably more engaging, exhibited heterogeneous results. High therapist interaction significantly improved the performance of the limb with the highest muscular strength for the hip during the swing phase, whereas the effects on the limb with the lowest muscular strength were less clear.