Gait parameters in school going children using a marker-less approach

The burden of course work in Indian schools has exposed the school children to various postural/gait disorders due to heavy backpack. Therefore, it is paramount to develop a low cost, non-intrusive and reliable method for calculation of gait parameters. This study assessed the spatiotemporal parameters such as height of earlobe (HoE), stride length (SL) and stride width (SW) using the markerless sensor Kinect v2 and conventional techniques pursued in Indian clinics. Sixty school children (aged 11 to 15 years) were monitored through both the techniques while performing walking trials. To assess the agreement between the techniques Bland-Altman 95% bias, percentage error (PE), Pearson's correlation coefficients (r(1)) and concordance correlation coefficients (r(2)) were determined. Each parameter obtained from both techniques possessed strong correlation (r(1) (and 2) > 0.90). Gait analysis using the Kinect V2 sensor is an acceptable, unobtrusive and economical method. The effect of relative backpack weight (RBW), i.e. (bag weight to body weight percentage) and strategies of backpack packing recommended by the American Occupational Therapy Association on the selected parameters was studied. The effect of RBW on the variation in parameters was evaluated using the regression curve whereas the effect of proper packing was evaluated by paired sample T test. RBW has positive correlation with SW (r(1) = 0.631), negative correlation with HoE (r(1) = -0.387) but shows no correlation with SL. Recommended packing strategy of schoolbag by AOTA shows results to reduce the unwanted variation in gait parameters.