We present a methodology based on computational neuromusculoskeletal models of the human body as a means of predicting the actions of muscles during dynamic motor tasks. In this scenario, experimental surface electromyograms (EMG) are used to “drive” the simulated muscles in the model. This also allows estimat ing muscle activation patterns for muscles from which EMGs cannot be measured and allows adjusting experimental EMG recording that may be subject to measurement errors. Furthermore, we present another methodology that uses a lowdimensional set of basic muscle activation primitives (APs) to model the resulting motor programs that coordinate the recruitment of muscles during human locomotion. The APs are then used to perform musculoskeletal simulation of locomotion tasks. We describe the theoretical aspects of the proposed methodology and discuss its implications in neurorehabilitation technologies. Furthermore, we present experimental results that demonstrate the benefits of the new method.
[1]
D. Lloyd,et al.
An EMG-driven musculoskeletal model to estimate muscle forces and knee joint moments in vivo.
,
2003,
Journal of biomechanics.
[2]
Massimo Sartori,et al.
Estimation of musculotendon kinematics in large musculoskeletal models using multidimensional B-splines.
,
2012,
Journal of biomechanics.
[3]
Enrico Pagello,et al.
Modeling the Human Knee for Assistive Technologies
,
2012,
IEEE Transactions on Biomedical Engineering.
[4]
M G Pandy,et al.
Static and dynamic optimization solutions for gait are practically equivalent.
,
2001,
Journal of biomechanics.
[5]
F. Lacquaniti,et al.
Five basic muscle activation patterns account for muscle activity during human locomotion
,
2004,
The Journal of physiology.