1994 Project Reports | Contents | Previous | Next | Home |
Pedaling Coordination: identification of muscle control strategies
Christine C Raasch, MS; Benjamin J Fregly, PhD; Felix E Zajac, PhD; Christine A Dairaghi, BS
Objective - The goal of this study is to better understand central nervous system coordination of muscles in the lower limb. Pedaling is a good task for this, since it is amenable to laboratory experimentation and modeling, yet complex enough that its control is interesting. Pedaling a stationary bike also does not involve the complicating factors of balance and posture. Still, while it may seem trivially easy to us because it is so well-learned, some neurologically-impaired patients have difficulty pedaling smoothly or quickly, and tend to rely heavily on the non-affected leg. We hope to gain some insight into how biomechanical factors (including gravity, mechanical constraints, muscle properties, etc.) and neuromuscular coordination influence pedaling. Computer models allow us to interpret experimental results and predict the consequences of different muscle control strategies, including pathological muscle groupings or synergies. This will help us in developing better assessment, intervention, and rehabilitative strategies.
Approach - Our model of the person pedaling (Figure 1) included the legs
and stationary bicycle (ergometer), and nine muscle groups on each leg. Muscle
excitations were found using an optimization algorithm which set the timing and
magnitude so that the simulation pedaled as fast as possible. For
constant-speed pedaling, the model was driven by net muscle joint torques
derived from measured data.
Figure 1. Dynamic
model of pedaling.
Experimental data was collected from healthy subjects pedaling a specially designed and instrumented ergometer. One group of subjects were asked to pedal at a smooth cadence with the ergometer emulating both normal load conditions and the higher inertial load characteristics of actual road bicycle riding. Other subjects pedaled from rest to maximal speed.
Conclusions - Analysis of a simulation that closely matched experimental data revealed that ankle and hip extensor muscles work together to deliver power to the crank during the downstroke. In contrast, our results indicate that the torque generated at the knee functions independently to propel crank through the top and bottom of the stroke, respectively, which prevents freewheel decoupling. During the upstroke, ankle extensor muscle activity aids in recovering the limb rather than propelling the crank.
Subject data showed changes in driving force to the crank and decreased cycle-to-cycle cadence variability when the inertial load was increased. We conclude that the neuromuscular control at higher inertial load may be simplified by the system's slower response to perturbations. That is, coordination must be finer at lower inertial loads to meet task demands. Thus, the use of gradually decreasing inertial loads may be a helpful strategy for training stroke patients to regain fine control of the affected limb.
Simulations of maximal-speed pedaling agreed with measured data from subjects performing the same task. Optimization results showed that different ankle muscle control strategies may be used; however, the calf muscles must be activated near the end of downstroke to prevent knee hyperextension. Although pedaling might be considered a simple, "push-pull" task, simulations with muscles excited in this synergistic manner did not produce smooth continuous motion. These results suggested that pedaling is more of a four-phase task, requiring timely deactivation of flexors/extensors followed by active propulsion through the top and bottom of stroke, provided mainly by two-joint muscles (e.g., hamstrings). Since some neurologically-impaired subjects exhibit problems similar to those seen in the simulation, we theorize that they may have difficulty breaking out of synergistic activation patterns.
Republished from the 1994 Rehabilitation R&D Center Progress Report. For
current information about this project, contact
Christine C Raasch.
