Weir Gillian, Alderson Jacqueline, Smailes Natalie, Elliott Bruce, Donnelly Cyril
Biomechanics Laboratory, The University Massachusetts, Amherst, United States.
School of Human Sciences, The University of Western Australia, Perth, Australia.
Int J Sports Med. 2019 Mar;40(3):191-199. doi: 10.1055/a-0756-9659. Epub 2019 Jan 10.
This study aimed to develop a 2-dimensional (2D) video screening tool capable of predicting an athlete's peak 3-dimensional (3D) knee moments during unplanned sidestepping. 2D video-based kinematic measures were simultaneously captured with 3D peak knee moments for 30 female field hockey players (15 junior, 15 senior). Intra- and intertester repeatability of 2D kinematic measures was performed. Then, linear regression models were used to model 3D knee moments from 2D kinematic variables utilizing 80% of the sample (n=24). Regression equations were then validated on the remaining 20% of the sample (n=6). Angular 2D measures had good-excellent intra- (ICC=0.936-0.998) and intertester (ICC=0.662-0.949) reliability. Displacement measures had poor-excellent intra- (ICC=0.377-0.539) and inter-tester (ICC=0.219-0.869) reliability. Significant independent predictors of peak knee moments were dynamic knee valgus, knee flexion angle at foot strike, trunk flexion range of motion (ROM), trunk lateral flexion, hip abduction and knee flexion ROM (P<0.05). Regression equations generated from these models effectively predicted peak knee extension, valgus and internal rotation moments (i. e., were not different from measured values P>0.05, ES<0.4) in the 20% subsample. 2D video-based measurements of an athlete's full body kinematics during unplanned sidestepping provide a reliable, specific, sensitive and cost-effective means for screening female team sport athletes.
本研究旨在开发一种二维（2D）视频筛查工具，该工具能够预测运动员在非计划侧步时的三维（3D）膝关节峰值力矩。对30名女子曲棍球运动员（15名青少年，15名成年人）同时采集基于2D视频的运动学测量数据和3D膝关节峰值力矩。对2D运动学测量进行了测试者内和测试者间的重复性研究。然后，利用样本的80%（n = 24），通过线性回归模型从2D运动学变量中模拟3D膝关节力矩。随后，在剩余20%的样本（n = 6）上对回归方程进行验证。角度2D测量具有良好至优秀的测试者内（ICC = 0.936 - 0.998）和测试者间（ICC = 0.662 - 0.949）可靠性。位移测量具有较差至优秀的测试者内（ICC = 0.377 - 0.539）和测试者间（ICC = 0.219 - 0.869）可靠性。膝关节峰值力矩的显著独立预测因素为动态膝外翻、足着地时的膝关节屈曲角度（ROM）、躯干前屈活动范围、躯干侧屈、髋关节外展和膝关节屈曲ROM（P < 0.05）。这些模型生成的回归方程在20%的子样本中有效预测了膝关节伸展、外翻和内旋峰值力矩（即与测量值无差异，P > 0.05，ES < 0.4）。在非计划侧步期间，基于2D视频的运动员全身运动学测量为筛选女子团体运动运动员提供了一种可靠、特异、敏感且经济高效的方法。
