Search

Abstract Post-stroke gait disorders involve altered lower limb kinematics. Recently, the endpoint of the lower limb has been used as a control variable to understand gait kinematics better. In a cross-sectional study of sixty-seven post-stroke patients, the limb extension angle and effective limb length during gait were used as input variables with a mixed Gaussian model-based probabilistic clustering approach to identify five distinct clusters. Each cluster had unique characteristics related to motor paralysis, spasticity, balance ability, and gait strategy. Cluster 1 exhibited high limb extension angle and length values, indicating increased spasticity. Cluster 2 had moderate extension angles and high limb lengths, indicating increased spasticity and reduced balance ability. Cluster 3 had low limb extension angles and high limb length, indicating reduced balance ability, more severe motor paralysis, and increased spasticity. Cluster 4 demonstrated high extension angles and short limb lengths, with a gait strategy that prioritized stride length in the component of gait speed. Cluster 5 had moderate extension angles and short limb lengths, with a gait strategy that prioritized cadence in the component of gait speed. These findings provide valuable insights into post-stroke gait impairment and can guide the development of personalized and effective rehabilitation strategies.