Details

Autor(en) / Beteiligte

• Kolke, Yasuharu
• Kawato, Mitsuo

Titel

Trajectory formation from surface EMG signals using a neural network model

Erscheinungsjahr

1993

Link zum Volltext

Quelle

IEEE Electronic Library (IEL)

Beschreibungen/Notizen

• In this study, multi-joint arm movements were estimated from surface EMG signals using an artificial neural network with a modular architecture. A forward dynamics model (FDM) of the human arm which takes into account non-linear muscle properties such as the length-tension and velocity-tension curves and complicated geometrical configurations of the musculoskeletal system, was constructed by training the artificial neural network. The FDM is especially useful for computational study of movement, and is crucial to smooth trajectory formation based on an optimal principle. It can also be used to elucidate various computational problems such as direct calculation of the virtual trajectories during movement, and may have biomedical applications such as functional electrical stimulation of paralyzed limbs.