Application of principal component analysis in clinical gait research: Identification of systematic differences between healthy and medial knee-osteoarthritic gait
Application of principal component analysis in clinical gait research: Identification of systematic differences between healthy and medial knee-osteoarthritic gait
复制标题
DOI：
10.1016/j.jbiomech.2013.06.032
复制
发表时间：
2013-09-03
影响因子：
2.4
通讯作者：
Andriacchi, T. P.
中科院分区：
文献类型：
--
作者：
Federolf, P. A.;Boyer, K. A.;Andriacchi, T. P.
为了成功地完成运动任务，运动控制系统必须遵守管理其部分协调的许多内部约束。本研究的目的是应用主成分（PC）分析，以检测健康受试者和内侧膝关节骨关节炎（OA）患者之间的节段协调的差异。假设（1）即使在小样本组中，也可以通过该方法识别全身运动模式的系统性差异，（2）这些差异将包括OA患者下半身和上身节段的代偿性运动。使用主成分分析法分析了5名健康人和5名OA参与者的三次步态试验的标记位置和地面反作用力。对于前10个PC向量确定PC分数的组差异，并且发现差异的那些PC向量的线性组合定义了判别向量。将原始试验投影到该判别向量上产生了显著的组间差异（t（d=8）=3.011; p = 0.017），膝关节OA患者的上身运动更大，与内外侧地面反作用力相关。这些结果有助于表征在相对较小的人群中全身步态模式对膝关节OA的适应性，并可能为开发干预措施以修改膝关节负荷提供更好的基础。基于PC的运动分析提供了一种高度灵敏的方法来识别与病理步态相关的特征性全身运动模式。(C)2013爱思唯尔有限公司保留所有权利。
For a successful completion of a movement task the motor control system has to observe a multitude of internal constraints that govern the coordination of its segments. The purpose of this study was to apply principal component (PC) analysis to detect differences in the segmental coordination between healthy subjects and patients with medial knee osteoarthritis (OA). It was hypothesized that (1) systematic differences in patterns of whole body movement would be identifiable with this method even in small sample sized groups and that (2) these differences will include compensatory movements in the OA patients in both the lower and upper body segments. Marker positions and ground reaction forces of three gait trials of 5 healthy and 5 OA participants with full body marker sets were analyzed using a principal component analysis. Group differences in the PC-scores were determined for the first 10 PC-vectors and a linear combination of those PC-vectors where differences were found defined a discriminant vector. Projecting the original trials onto this discriminant vector yielded significant group differences (t(d=8)=3.011; p = 0.017) with greater upper body movement in patients with knee OA that was correlated with the medial-lateral ground reaction force. These results help to characterize the adaptation of whole-body gait patterns to knee OA in a relatively small population and may provide an improved basis for the development of interventions to modify knee load. The PC-based motion analysis offered a highly sensitive approach to identify characteristic whole body patterns of movement associated with pathological gait. (C) 2013 Elsevier Ltd. All rights reserved.