两种不同肌肉力求解算法的对比分析

目的 探究静态优化(static optimization, SO)算法和计算肌肉控制(computed muscle control, CMC)算法求解肌肉力结果的异同，为研究人员选择合适的肌肉力求解算法和横向对比不同研究的结果提供依据。方法 以4种不同速度下跑步的步态作为研究动作，分别利用SO和CMC算法求解单个步态周期中下肢主要肌肉的肌肉力和肌肉激活，并进行对比分析。结果 在参与跑步的10块主要肌肉中，除了胫骨前肌和股直肌外，两种算法求解的肌肉力和肌肉激活具有相似的变化曲线，且相关系数均在0.91以上，但SO算法求解的肌肉力峰值偏高，肌肉激活峰值的位置约有10 ms延迟。结论 在动作分析时，如果研究重点在肌肉发力的时序以及不同肌肉之间的贡献比时，两种算法的选择没有太大区别，建议选择简单高效的SO算法。在横向对比不同研究中SO和CMC算法求解的肌肉力和肌肉激活结果时，需要考虑两者的差异。

Comparative Analysis on Two Different Algorithms for Estimating Muscle Forces

Objective To explore the differences between the result of static optimization (SO) and computational muscle control (CMC) algorithms for estimating muscle forces, so as to provide references for researchers to choose the appropriate algorithm and make horizontal comparison of the results from different studies. Methods Targeting at a single gait cycle running at four different speeds, SO and CMC algorithms were used to calculate forces and activations of the major muscles in lower limbs, and the results were compared and analyzed. Results Among the 10 major muscles participating in running, except for anterior tibial and rectus femoris, muscle forces and muscle activations solved by the two algorithms had similar curves with correlation coefficients more than 0.91, and the peak value of muscle forces solved by SO was higher and the positions of peak muscle activation had a 10 ms delay. Conclusions In movement analysis, if the research focuses on the timing of muscle forces and the contribution ratios among different muscles, there is not too big difference and SO algorithm is recommended for its simplicity and efficiency. For horizontal comparison of muscle forces and muscle activations estimated by SO and CMC algorithms in different studies, the differences between the two algorithms should be considered.