ABSTRACT
Human locomotion is an important activity of daily life. Every individual human walk with a specific gait pattern, and there underlay a neural command mechanism that works in synergy with that particular walking pattern. The presented study explored the neural command mechanism for walking at different speeds using muscle synergies and proposed a walking speed prediction algorithm. The difference in the neural control mechanism has been examined by comparing muscle synergies extracted for different walking speeds. The linear factorization technique was implemented to extract synergies from four lower limb muscle’s electromyogram signals, recorded at different walking speeds (3–11 km/hr). A 2- synergy model was found most suitable for all walking conditions; each synergy served a particular purpose during a complete gait cycle. The activation pattern of muscle synergy has shown a significant difference in terms of duration, timing, and magnitude (p-value < 0.05). Further, the same synergy model has been utilized for walking speed prediction, and an average accuracy of 97.28 ± 0.43% has been achieved. The performance of the developed prediction module has also been estimated over the lower range of walking speed (0.5–2.5 km/hr). The time complexity of the proposed algorithm has been found within the boundaries of its real-time implementation. The outcomes of the research work will be helpful for lower limb assistive/rehabilitative device control and the development of an effective human-machine interface.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author(s).
Additional information
Notes on contributors
Rohit Gupta
Rohit Gupta completed his PhD from Thapar Institute of Engineering and Technology, Patiala. Currently, he is post-doctoral fellow at Indian Institute of Technology Delhi.
Ravinder Agarwal
Ravinder Agarwal is currently professor in Electrical and Instrumentation Engineering Department at Thapar University, Patiala. He has published more than 125 research papers in reviewed international journals of repute. Till now, he has guided 20 PhD students. Email: [email protected]