OpenSim 把 EMG 当成参数来更好地估计肌腱的动态参数，并不是直接生成 EMG 肌电信号。Fuglevand 的模型被广泛用于肌肉力，运动神经元活动，和肌电信号的研究，但是受限于等长收缩。
Reasons for the lack of an integrated and precise EMG simulator feasible for voluntary movements:
- the absence of models that link EMG generation to movement biomechanics
- current advanced EMG models are not efficient enough to adapt to the non-stationary physiological parameters during voluntary movements
上述两个问题被之前所提出的模型 BioMime 解决。
Ma S, Clarke AK, Maksymenko K, Deslauriers-Gauthier S, Sheng X, Zhu X, et al. Conditional generative models for simulation of EMG during naturalistic movements. arXiv:2211.01856 [preprint], 2022. Available from https://arxiv.org/abs/2211.01856.
BioMime：
- 输入：physilogical parameters
- 输出：dynamic MUAP signals
NeuroMotion 包含 3 个模块：
- MSK: define and visualise the movement and estimating the muscle fibre lengths and muscle activations during the movement.
- BioMime: utilise muscle fibre lengths to simulate the dynamic MUAPs during the movement.
- Motor unit pool: receives the neural inputs derived from the muscle activations and outputs stimulations to each muscle in the format of spike trains.
Toolbox functions
- define the movement of the ARMs model (Section Define movement)
- track the changes in physiological parameters (Section Estimate parameter changes)
- set common drives to motoneuron pools (Section Set neural inputs to motoneuron pools)
- define the structure of motor unit pools (Section Configure motor unit pools)