Evidence for generalized motor programs using gait pattern analysis.

Human intra limb gait kinematics were analyzed via statistical and structural pattern recognition methods to determine the role of relative timing of limb segments within and between modes of gait. Five experienced runners were filmed while walking (3-6 km/hour) and running (8-12 km/hour) on a motor driven treadmill. Kinematic data consisted of relative timing of the four phases of the Philippson step cycle and intersegmental limb trajectories, determined from angle-angle diagrams. Despite marked decreases in absolute time durations within gaits remained constant over the speeds which were studied. Although a 2-fold increase in locomotor speed occurred in walking and a 1.5-fold speed increase occurred within running, the percentage of time spent in each of the Philippson phases was not significantly changed. However, significant differences in the time percentages and sequences of the step cycle phases were found between walking and running. Correlations between limb segment trajectories occurring in the different gaits showed strong coherence for overall step cycle patterns, but within step cycle phases and across speeds, selective phases displayed little correspondence.