利用位移資訊評估特定對象肌肉模型參數之方法

A preliminary study on estimating muscle model parameters of a specific subject using displacement data

簡昱凡(Yu-Fan Chien)

指導教授 ： 詹魁元

國立臺灣大學/工學院/機械工程學研究所/碩士(2022年)

https://doi.org/10.6342/NTU202203007

摘要

生物體的組成複雜，加上內部生理數據難以量測，實務上生物力學模型提供了由外部量測估計內部數據的管道，故無論在工業研發或醫療研究都佔有具足輕重的地位。而在模型中表現出生物的個體差異，對於講求客製化的工業4.0與追求個人化精準醫療的時代，是必要卻充滿挑戰的任務。有鑑於當前針對個人化的肌肉模型參數調整方法，仍受到儀器成本與量測場域之限制，本研究針對突破這些限制的可行方法進行探討，提出一個僅利用位移量測資訊，對模型肌肉參數進行評估之流程。本研究以自行生成的模擬軌跡取代量測，接著透過軌跡最佳化估測肌肉骨骼模型動態，使不同參數之模型重現生成之目標動作，並根據此重現結果與目標的差異，判斷該參數組合是否與欲校正之目標相同。此外，本研究也探討了具多組目標動作的評估，可更精準篩選出能同時滿足多組動作之肌肉參數，同時也歸納出有利於提升參數評估效率的動作應具備之特性，包含主作用肌群的交替，以及運動方向的改變。最後，本研究以簡易上臂模型作為案例，證實了前述評估準則的可行性，並為未來可能的延伸應用提供了可用於非模擬環境的驗證方式。

關鍵字

個人化模型 ； 肌肉肌腱參數 ； 肌肉動態模擬 ； 希爾肌肉模型 ； 參數評估

並列摘要

By providing the connection between noninvasive measurement and estimation of undersurface data, biomechanics simulations play an essential role in both industry and the medical field. However, challenges emerge between individuals in pursuing customizing and personalization treatments. Subject-specific muscle model parameters estimation methods are still limited in the overall cost and the environment. Therefore, in this work we present a workflow of muscle model parameters estimation using only displacement measurement data that has both cost and applicating advantages. We first estimate the dynamics of the musculoskeletal model by trajectory optimization with the measurement data. Since this study aims to validate the method, the measurement data is temporarily obtained via model simulations that are treated as in the field tests with parameters to be estimated. The different parameters samples are then evaluated by the result of
the dynamics simulation to see if they are potentially matching the goal subject.
In addition, the number of motions used in parameter estimation is proved to be contributive for eliminating improper parameters. Also, criteria for the motions used in estimating process were built, which are the change of main active muscle and moving direction. Finally, we validated the conclusions above with a simple model of an arm and provided a validating method that is feasible in a real-world experimental environment for possible future applications.

並列關鍵字

subject-specific model ； muscle-tendon parameters ； muscle dynamics simulation ； Hill muscle model ； parameter estimation

參考文獻

[1] S. R. Dubowsky, J. Rasmussen, S. A. Sisto, and N. A. Langrana, “Validation of a musculoskeletal model of wheelchair propulsion and its application to minimizing shoulder joint forces,” Journal of Biomechanics, vol. 41, no. 14, pp. 2981–2988, 2008.

[2] E. P. Grabke, K. Masani, and J. Andrysek, “Lower limb assistive device design optimization using musculoskeletal modeling: a review,” Journal of Medical Devices, vol. 13, no. 4, 2019.

[3] M. G. Pandy and T. P. Andriacchi, “Muscle and joint function in human locomotion,” Annual Review of Biomedical Engineering, vol. 12, pp. 401–433, 2010.

[4] L. L. Presti and M. La Cascia, “3d skeleton-based human action classification: A survey,” Pattern Recognition, vol. 53, pp. 130–147, 2016.

[5] G. Monnier, F. Renard, A. Chameroy, X. Wang, and J. Trasbot, “A motion simulation approach integrated into a design engineering process,” SAE Transactions, pp. 1118– 1123, 2006.