Abstract：
(1) Background: Significant advances in digital modelling worldwide have been attributed to the practical application of digital musculoskeletal (MS) models in clinical practice. However, the vast majority of MS models are designed to assess adults\' mobility, and the range suitable for children is very limited. (2) Methods: Seventeen healthy and 4 cerebral palsy (CP) children were recruited for the gait measurements. Surface electromyography (EMG) and ground reaction forces (GRFs) were acquired simultaneously. The MS model of the adult was adapted to the child and simulated in AnyBody. The differences between measured and MS model-estimated GRFs and muscle activations were evaluated using the following methods: the root-mean-square error (RMSE); the Pearson coefficient r; statistical parametric mapping (SPM) analysis; the coincidence of muscle activity. (3) Results: For muscle activity, the RMSE ranged from 10.4% to 35.3%, the mismatch varied between 16.4% and 30.5%, and the coincidence ranged between 50.7% and 68.4%; the obtained strong or very strong correlations between the measured and model-calculated GRFs, with RMSE values in the y and z axes ranged from 7.1% to 17.5%. (4) Conclusions: Child-adapted MS model calculated muscle activations and GRFs with sufficient accuracy, so it is suitable for practical use in both healthy children and children with limited mobility.
摘要：
（1）背景：全球数字建模的重大进展归因于数字肌肉骨骼（MS）模型在临床实践中的实际应用。然而,绝大多数MS模型旨在评估成年人的流动性，适合儿童的范围非常有限。（2）方法：招募17名健康儿童和4名脑瘫（CP）儿童进行步态测量。同时获得表面肌电图（EMG）和地面反作用力（GRF）。成人的MS模型适用于儿童，并在AnyBody中进行了模拟。使用以下方法评估测量的和MS模型估计的GRF和肌肉激活之间的差异：均方根误差（RMSE）;皮尔逊系数r;统计参数映射（SPM）分析;肌肉活动的重合。（3）结果：对于肌肉活动，RMSE范围从10.4%到35.3%，不匹配在16.4%到30.5%之间，和一致性范围在50.7%和68.4%之间;所获得的强或非常强的相关性之间的测量和模型计算的GRFs,y轴和z轴的RMSE值范围为7.1%至17.5%。(4)结论:儿童适应MS模型计算肌肉激活和GRFs具有足够的准确性,因此，它适用于健康儿童和行动不便儿童的实际使用。
