Transverse-contour modeling of trunk muscle-distributed forces and spinal loads during lifting and twisting.

Abstract

An electromyography-assisted biomechanical model was developed using electromyographic (surface and in-dwelling) data collected during asymmetric lifting and twisting activities. To develop a biomechanical model of the lumbar region that considers the ability of the broad, flat muscles of the trunk (external obliques, internal obliques and latissimus dorsi) to activate different anatomic regions at different intensity levels and then uses this information to describe the spine reaction forces that result during lifting and twisting tasks. Many biomechanical models of the lumbar region use single-vector representations for the external oblique, internal oblique, and latissimus dorsi muscles. This simplification limits the description of the complexity of the resultant forces produced by these muscles and does not consider their differential activation capacity. Human subjects performed lifting and twisting exertions while muscle electromyographic activities were sampled at one location on the rectus abdominis and erector spinae muscles and at multiple locations on the latissimus dorsi, external oblique, and internal oblique muscles. These data were used in conjunction with in vivo digitized muscle origin and insertion points to predict muscle forces and spine loads through the use of the electromyography-assisted modeling method. The measures of model performance such as percentage of error (6-21%) in the prediction of the external torques, correlations (0.83-0.98) between internal and external torques and the values of predicted muscle force capacity were all similar to data collected in previous electromyography-assisted models, but the predictions of spinal loading, particularly shear forces, were quite different. The results have shown that by modeling the broad, flat muscles of the torso using multiple-force vectors, the calculated shear forces in the spine were reduced. The multivector, transverse-contour model developed in this research illustrates the importance of realistic multiple-vector modeling and the importance of considering the selective-activation capacity of the abdominal oblique musculature.