Martínez Fabio, Gómez Francisco, Romero Eduardo
Bioingenium Research Group - School of Medicine, National University of Colombia, Bogotá, DC, Colombia.
Comput Methods Biomech Biomed Engin. 2011 Jun;14(6):561-72. doi: 10.1080/10255842.2010.486761.
The gait pattern of a particular patient can be altered in a large set of pathologies. Tracking the body centre-of-mass (CoM) during the gait allows a quantitative evaluation of these diseases at comparing the gait with normal patterns. A correct estimation of this variable is still an open question because of its non-linearity and inaccurate location. This paper presents a novel strategy for tracking the CoM, using a biomechanical gait model whose parameters are determined by a Bayesian strategy. A particle filter is herein implemented for predicting the model parameters from a set of markers located at the sacral zone. The present approach is compared with other conventional tracking methods and decreases the calculated root mean squared error in about a 56% in the x-axis and 59% in the y-axis.
特定患者的步态模式在大量病症中会发生改变。在步态过程中追踪身体质心（CoM），通过将步态与正常模式进行比较，能够对这些疾病进行定量评估。由于该变量的非线性和位置不准确，对其进行准确估计仍是一个悬而未决的问题。本文提出了一种追踪CoM的新策略，使用一种生物力学步态模型，其参数由贝叶斯策略确定。本文采用粒子滤波器从位于骶骨区域的一组标记物预测模型参数。将本方法与其他传统追踪方法进行比较，在x轴上计算出的均方根误差降低了约56%，在y轴上降低了59%。
