DocumentCode :
3181683
Title :
Human joint motion estimation for electromyography (EMG)-based dynamic motion control
Author :
Qin Zhang ; Hosoda, Ryo ; Venture, G.
Author_Institution :
State Key Lab. of Digital Manuf. Equip. & Technol., Huazhong Univ. of Sci. & Technol., Wuhan, China
fYear :
2013
fDate :
3-7 July 2013
Firstpage :
21
Lastpage :
24
Abstract :
This study aims to investigate a joint motion estimation method from Electromyography (EMG) signals during dynamic movement. In most EMG-based humanoid or prosthetics control systems, EMG features were directly or indirectly used to trigger intended motions. However, both physiological and nonphysiological factors can influence EMG characteristics during dynamic movements, resulting in subject-specific, non-stationary and crosstalk problems. Particularly, when motion velocity and/or joint torque are not constrained, joint motion estimation from EMG signals are more challenging. In this paper, we propose a joint motion estimation method based on muscle activation recorded from a pair of agonist and antagonist muscles of the joint. A linear state-space model with multi input single output is proposed to map the muscle activity to joint motion. An adaptive estimation method is proposed to train the model. The estimation performance is evaluated in performing a single elbow flexion-extension movement in two subjects. All the results in two subjects at two load levels indicate the feasibility and suitability of the proposed method in joint motion estimation. The estimation root-mean-square error is within 8.3% ~10.6%, which is lower than that being reported in several previous studies. Moreover, this method is able to overcome subject-specific problem and compensate non-stationary EMG properties.
Keywords :
adaptive estimation; biomechanics; crosstalk; electromyography; estimation theory; mean square error methods; medical control systems; medical signal processing; motion estimation; prosthetics; EMG signals; EMG-based dynamic motion control; EMG-based humanoid control systems; EMG-based prosthetics control systems; adaptive estimation method; antagonist muscles; dynamic movements; electromyography based dynamic motion control; human joint motion estimation method; joint torque; linear state-space model; motion velocity; multiinput single output; muscle activation; muscle activity map; nonstationary EMG properties; root-mean-square error method; single elbow flexion-extension movement; subject-specific problem; trigger intended motions; Angular velocity; Dynamics; Elbow; Electromyography; Estimation; Joints; Muscles;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Engineering in Medicine and Biology Society (EMBC), 2013 35th Annual International Conference of the IEEE
Conference_Location :
Osaka
ISSN :
1557-170X
Type :
conf
DOI :
10.1109/EMBC.2013.6609427
Filename :
6609427
Link To Document :
بازگشت