Using a Robotic Device to Quantify the Impact of Conventional Gait Training on Ankle Proprioception After Stroke

Ankle proprioception is important for gait control and is often impaired after stroke. It is currently unclear if conventional gait training can improve ankle proprioception. Here we used a novel robotic device, Ankle Measuring Proprioceptive Device (AMPD), to quantify key aspects of ankle proprioception before and after four weeks of gait training delivered to 24 individuals who were in the chronic phase of stroke. We implemented two assessments with AMPD to quantify passive and active and dynamic/static proprioception. For the Crisscross assessment, AMPD moved the ankles past each other and the participants pushed a button to indicate when the ankles passed each other, thereby testing passive, dynamic proprioception. For the Joint Position Reproduction (JPR) assessment, AMPD moved the impaired ankle in a random pattern and the participant tracked with their unimpaired ankle, thereby testing active, dynamic proprioception. We also incorporated static periods into JPR to quantify static position sense. At baseline, active and passive ankle proprioception errors, as well as static and dynamic errors, were significantly increased compared to age-matched controls and were moderately correlated. Gait training improved only static ankle position sense, with greater improvement for more severely impaired individuals. These findings suggest that stroke damages multiple aspects of ankle proprioception in an interrelated manner. In contrast, conventional gait training improved only static ankle proprioception ability. This suggests a need for an array of training methodologies to comprehensively retrain ankle proprioception.