Longitudinal Investigation of Postural Strategies Associated with Improvement in Gait Speed Following Adult Spinal Deformity Surgery using Machine Learning.
retrospective_cohort · Level III
Where this comes from
- Record sourced from PubMed, PMID 41860195.
- Also identified by DOI 10.1097/BRS.0000000000005691.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
Retrospective longitudinal cohort study. To identify postural strategies associated with improvement in gait speed following spinal fixation from thoracic vertebrae to pelvis using machine learning. Although the spinal fixation for adult spinal deformity (ASD) is effective in improving radiographic parameters and quality of life, improvement in gait function is more challenging. The underlying reason has not been elucidated due to the lack of longitudinal gait analysis research focusing on change in speed and posture. Three-dimensional gait analysis was performed before and 6 months following the surgery to investigate the gait speed and anteroposterior distance from center of mass to center of pressure in trailing limb (COM-COP distance). The relationship between COM-COP distance and gait speed were demonstrated. Postoperative gait patterns were categorized through k-means clustering, and the disease-specific gait characteristics associated with COM-COP distance were extracted. Eighty-nine patients (mean age: 73.4 years) with ASD were included. Changes in gait speed between pre- and post-surgery were associated with changes in COM-COP distance (r=0.607, P<0.001). Postoperative gait speed was associated with COM-COP distance (r=0.873, P<0.001). The COM-COP distance comprised the hip sagittal angle, power, and ankle plantarflexion moment (R2=0.793, P<0.001). Two clusters A (n=37) and B (n=52) were identified. The postoperative COM-COP distance was significantly associated with hip sagittal power on terminal stance (r=-0.487, P<0.001) in the younger cluster A and maximum ankle plantarflexion moment (r=0.576, r<0.001) in the older cluster B. Cluster A significantly improved gait speed but cluster B did not improve gait speed and COM-COP distance following the surgery. The COM-COP distance composed of hip and ankle movement was associated with the improvement in gait speed. These findings imply the new rehabilitation strategies to improve gait speed. Ⅲ.