The goal of this research was to develop new stimulation patterns for functional electrically stimulated (FES) recumbent pedaling that would enable an individual with a spinal cord injury (SCI) to exercise for a longer period of time and at a higher workrate while involving more leg muscles and thereby obtain greater exercise benefit from FES pedaling exercise than is possible with existing ergometers. These goals were addressed in three phases using both theoretical analyses and experiments. The objectives of the first phase were to determine the function and mechanical contributions of the individual muscles to driving the limbs and crank and how the muscle function and mechanical work contributions varied with changes in pedaling rate. This information was useful in determining the influence of pedaling rate on the subsequent FES pedaling studies. To satisfy the objectives, experimental electromyography (EMG), kinematic, and kinetic data were collected from 16 subjects as they pedaled a recumbent ergometer at three pedaling rates (40, 50, and 60 rpm) and a single workrate (50 W). The experimental data were analyzed to determine the muscle excitation on and off timing, the intersegmental moments of the hip, knee, and ankle, right pedal angle, crank torque, and normal and tangential pedal forces. These data patterns were used in conjunction with a forward dynamic model to develop computer simulations of recumbent pedaling. For each pedaling rate, muscle power was analyzed to determine the influence of individual muscle forces on the accelerations of the limbs and crank, which in turn permitted the calculation of individual muscle mechanical energy contributions to recumbent pedaling. The uniarticular hip and knee extensors generated 65 percent of the total mechanical work in recumbent pedaling. The triceps surae muscles transferred power from the limb segments to the crank and the bi-articular muscles that crossed the hip and knee delivered power to the crank during the leg transitions between flexion and extension. The functions of the individual muscles did not change with pedaling rate, but the mechanical energy generated by the knee extensors and hip flexors decreased as pedaling rate increased. In was concluded that it is possible to manipulate the individual muscle power contributions to the crank and limb segments in recumbent pedaling by varying the pedaling rate.

The objective of the second phase of the study was to use a forward dynamic simulation of FES pedaling by SCI individuals to determine muscle excitation timing patterns for three upper leg muscle sets (Stim3) and three upper leg and two lower leg muscle sets (Stim5) that lead to increased endurance and work output in FES pedaling by SCI individuals. To satisfy these objectives a forward dynamic simulation of FES pedaling was developed to determine electrical stimulation on and off times that minimize the muscle stress-time integral of the stimulated muscles. The computed electrical stimulation on and off times differed from those utilized by a commercially available FES ergometer (StimErg) and resulted in a 17 and 11 percent decrease in the muscle stress-time integral for the three upper leg muscle sets and five upper and lower leg muscle sets, respectively. The mechanical energy generated by the hamstrings increased by 20 percent over a crank cycle compared to the existing FES ergometer stimulation timing patterns. The simulation results indicated that the lower leg muscle sets did not generate sufficient mechanical energy to reduce the energy contributions of the upper leg muscle sets. Based on the simulation results it was concluded that the computed stimulation on and off times had the potential to prolong pedaling and thereby provide improved cardiorespiratory and muscle training outcomes for individuals with spinal cord injury.

The objective of the third phase of the study was to verify the theoretical results from the forward dynamic simulations experimentally. Mechanical work, rate of oxygen consumption (V0₂), and blood lactate data were measured from SCI subjects (injury level T4-T12) as they pedaled using the existing FES ergometer stimulation timing patterns (StimErg), the computed stimulation timing patterns for the upper leg muscle sets (Stim3), and the upper and lower leg muscle sets (Stim5) on repeated trials. On average, the subjects performed significantly more work, 11 percent (p<0.05), pedaling with Stim3 than StimErg prior to the termination criterion. The average rate of oxygen consumption and blood lactate concentration associated with StimErg were not significantly different from the corresponding averages for Stim3. Neither the work nor the rate of oxygen consumption associated with Stim3 was significantly different from the corresponding averages for Stim5. However, the blood lactate concentration associated with Stim3 was significantly lower than that for Stim5 (p<0.05). The results indicated that electrical stimulation timing patterns that minimize the stress-time integral enabled 75 percent of the SCI subjects to perform more work than existing FES ergometer electrical stimulation patterns but do not significantly affect the metabolic response to the exercise when only the upper leg muscle sets are activated. In addition, electrical stimulation timing patterns that incorporate the lower leg muscles do not affect either the work performed or the rate of oxygen consumption but do increase the blood lactate concentrations. The increased mechanical work performed with Stim3 provides support for minimizing the stress-time integral as a means to prolong the endurance of FES pedaling. Incorporating the lower leg muscles in the exercise did not decrease the work accomplished but did involve more muscles which is a benefit to SCI individuals Finally, the experimental results supported the outcomes from the forward dynamic simulation of FES pedaling by SCI individuals.