Modified electromyography-assisted optimization approach for predicting lumbar spine loading while walking with backpack loads.

Abstract

This study modified an electromyography-assisted optimization approach for predicting lumbar spine loading while walking with backpack loads. The modified-electromyography-assisted optimization approach eliminated the electromyography measurement at maximal voluntary contraction and adopted a linear electromyography-force relationship. Moreover, an optimal lower boundary condition for muscle gain was introduced to constrain the trunk muscle co-activation. Anthropometric information of 10 healthy young men as well as their kinematic, kinetic, and electromyography data obtained while walking with backpack loads were used as inputs in this study. A computational algorithm was used to find and analyse the sensitivity of the optimal lower boundary condition for achieving minimum deviation of the modified-electromyography-assisted optimization approach from the electromyography-assisted optimization approach for predicting lumbosacral joint compression force. Results validated that the modified-electromyography-assisted optimization approach (at optimal lower boundary condition of 0.92) predicted on average, a non-significant deviation in peak lumbosacral joint compression force of -18 N, a standard error of 9 N, and a root mean square difference in force profile of 73.8 N. The modified-electromyography-assisted optimization approach simplified the experimental process by eliminating the electromyography measurement at maximal voluntary contraction and provided comparable estimations for lumbosacral joint compression force that is also applicable to patients or individuals having difficulty in performing the maximal voluntary contraction activity.