Chuang Li

The goal of this study was to examine the optimal strategies for the recognition of gait phase based on surface electromyogram (sEMG) of leg muscles while children with cerebral palsy (CP) walked on a treadmill. Ten children with CP were recruited to participate in this study. sEMG from eight leg muscles and leg position signals were recorded while subjects walked on a treadmill. The position signals of left and right legs were used to develop a five gait sub-phases classifier, i.e., mid stance, terminal stance, pre-swing, mid swing, and terminal swing. Seven feature sets of sEMG signals were tested in recognizing the five gait sub-phases of children with CP. Results from this study indicated that the recognition performance of mean absolute value and zero crossing was better than that with other feature sets when using support vector machine (average classification accuracy was 89.40%). Further, we found that the performance of gait phase recognition is relatively better in pre-swing than other sub-phases, and the performance of gait phase recognition is relatively poorer in mid-swing than other sub-phases. Results from this study may be used to develop an intention-driven robotic gait training system/paradigm for assisting walking in children with CP through robotic training.

Near-infrared spectroscopy is an efficient, low-cost technology that has potential as an accurate method in detecting the nitrogen content of natural rubber leaves. Successive projections algorithm (SPA) is a widely used variable selection method for multivariate calibration, which uses projection operations to select a variable subset with minimum multi-collinearity. However, due to the fluctuation of correlation between variables, high collinearity may still exist in non-adjacent variables of subset obtained by basic SPA. Based on analysis to the correlation matrix of the spectra data, this paper proposed a correlation-based SPA (CB-SPA) to apply the successive projections algorithm in regions with consistent correlation. The result shows that CB-SPA can select variable subsets with more valuable variables and less multi-collinearity. Meanwhile, models established by the CB-SPA subset outperform basic SPA subsets in predicting nitrogen content in terms of both cross-validation and external prediction. Moreover, CB-SPA is assured to be more efficient, for the time cost in its selection procedure is one-twelfth that of the basic SPA.