Sensor-Derived Trunk Stability and Gait Recovery: Evidence of Neuromechanical Associations Following Intensive Robotic Rehabilitation

Abstract

This quantitative observational study with pre-post design aimed to examine joint-specific kinematic adaptations and the relationship between trunk stability and spatiotemporal gait parameters following intensive robotic rehabilitation. A total of 12 neurological patients completed 16 sessions of gait training using the Tecnobody Smart Gravity Walker. Pre- and post-training kinematic data were collected for bilateral hip and knee flexion-extension, trunk flexion-extension, trunk lateral flexion, and center-of-gravity displacement. Waveforms were normalized to 100% stride. Paired t-tests assessed pre-post differences, and correlations examined associations between trunk stability and gait performance. Significant increases were found in right hip flexion-extension (t = 3.44, p < 0.001), trunk flexion-extension (t = 9.49, p < 0.001), and center-of-gravity displacement (t = 15.15, p < 0.001), with reduced trunk lateral flexion (t = -8.64, p < 0.001). Trunk flexion-extension correlated with gait speed (r = 0.74), step length (r = 0.68), and stride length (r = 0.71); trunk lateral flexion correlated with cadence (r = 0.66) and stride length (r = 0.70). Intensive robotic rehabilitation improved trunk and hip kinematics, supporting trunk stability as an important biomechanical correlate of gait recovery. Sensor-derived metrics revealed strong neuromechanical coupling between postural control and locomotion in neurological patients.