Associations between gait patterns, brain lesion factors and functional recovery in stroke patients

Brain CT scans and neurological condition were evaluated in 74 stroke patients. Firstly, we found that using a classification-tree technique based on CT scan parameters (an innovative method, analyzing four parameters simultaneously) coincided with our previously proposed kinematic artificial neural network (ANN) classification technique for 71.3% of patients. Lesion size and location were found to be the most significant CT scan predictors of gait classification. Secondly, we sought to gauge post-rehabilitation functional recovery in patients within the same three groups of gait pattern. We found significant differences in scores between the three gait pattern groups, before and after rehabilitation (Kruskal-Wallis test, p<0.001), while significant improvement was observed in each group (Wilcoxon text; p<0.01). We conclude that patient classification into pathological gait groups on the basis of gait or CT scan parameters may serve as an early predictor of future functional outcome.