Loading Recognition for Lumbar Exoskeleton Based on Multi-Channel Surface Electromyography from Low Back Muscles

Jiang，Naifu1; Wang，Lin1; Wang，Dashuai2; Fang，Peng1; Wu，Xinyu1; Li，Guanglin1

2024

10.1109/TBME.2024.3363212

IEEE Transactions on Biomedical Engineering   影响因子和分区

0018-9294

1558-2531

Lumbar exoskeleton is an assistive robot, which can reduce the risk of injury and pain in low back muscles when lifting heavy objects. An important challenge it faces involves enhancing assistance with minimal muscle energy consumption. One of the viable solutions is to adjust the force or torque of assistance in response to changes in the load on the low back muscles. It requires accurate loading recognition, which has yet to yield satisfactory outcomes due to the limitations of available measurement tools and load classification methods. This study aimed to precisely identify muscle loading using a multi-channel surface electromyographic (sEMG) electrode array on the low back muscles, combined with a participant-specific load classification method. Ten healthy participants performed a stoop lifting task with objects of varying weights, while sEMG data was collected from the low back muscles using a 3x7 electrode array. Nineteen time segments of the lifting phase were identified, and time-domain sEMG features were extracted from each segment. Participant-specific classifiers were built using four classification algorithms to determine the object weight in each time segment, and the classification performance was evaluated using a 5-fold cross-validation method. The artificial neural network classifier achieved an impressive accuracy of up to 96%, consistently improving as the lifting phase progressed, peaking towards the end of the lifting movement. This study successfully achieves accurate recognition of load on low back muscles during the object lifting task. The obtained results hold significant potential in effectively reducing muscle energy consumption when wearing a lumbar exoskeleton.

Atmospheric measurements Biomedical measurement Electrodes Exoskeleton Exoskeletons Hip lift assistance load classification low back pain Muscles Particle measurements surface electromyography

SCI ; EI

英语

其他

20240715545363

Electrodes ; Electromyography ; Energy utilization ; Exoskeleton (Robotics) ; Loads (forces) ; Neural networks ; Time domain analysis

Structural Design:408 ; Biological Materials and Tissue Engineering:461.2 ; Energy Utilization:525.3 ; Robotics:731.5 ; Mathematics:921

ENGINEERING

2-s2.0-85184795817

Scopus

1.CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems, Shenzhen Institute of Advanced Technology (SIAT), Chinese Academy of Sciences (CAS), and the SIAT Branch, Shenzhen Institute of Artificial Intelligence and Robotics for Society, Shenzhen, China
2.School of Microelectronics, Southern University of Science and Technology, Shenzhen, China

Jiang，Naifu,Wang，Lin,Wang，Dashuai,et al. Loading Recognition for Lumbar Exoskeleton Based on Multi-Channel Surface Electromyography from Low Back Muscles[J]. IEEE Transactions on Biomedical Engineering,2024,PP(99):1-9.

Jiang，Naifu,Wang，Lin,Wang，Dashuai,Fang，Peng,Wu，Xinyu,&Li，Guanglin.(2024).Loading Recognition for Lumbar Exoskeleton Based on Multi-Channel Surface Electromyography from Low Back Muscles.IEEE Transactions on Biomedical Engineering,PP(99),1-9.

Jiang，Naifu,et al."Loading Recognition for Lumbar Exoskeleton Based on Multi-Channel Surface Electromyography from Low Back Muscles".IEEE Transactions on Biomedical Engineering PP.99(2024):1-9.