A method for quantification of directional patterns of muscle activation at the lower limb

Several studies have examined muscle activation patterns during movements or static torques exerted in different directions in human subjects as well as in trained monkeys. However, no statistical approach has ever been proposed to characterize directional patterns of muscle activation in a given population. This report describes a method for the quantification and statistical analysis of directional patterns of muscle activation at the lower limb. This method was used with a group of 18 healthy subjects. Using a multidirectional dynamometer for the lower limb, subjects were asked to exert static torques (at approximately 5 and 15% of the maximal voluntary contraction) in eight directions covering a 360 degrees range at the hip. These were hip abduction, adduction, flexion, extension, and the intermediate directions (e.g., hip flexion and abduction). Electromyographic (EMG) activities of eight lower limb muscles were recorded during these efforts using surface electrodes. For each muscle, the mean rectified EMG activity obtained in each direction was considered a vector. Then, the overall directional pattern of activation for a given muscle was quantified for each subject by a vectorial summation of these vectors. The angle of the resultant vector was used along with the other subjects' angular values to calculate a mean vector representing the muscle's directional pattern of activation for the complete sample. This mean vector was used to perform the statistical analysis of the data. More specifically, the Rayleigh test for circular data was performed at both torque levels to determine, through the length of the mean vector, if a muscle showed a directional specificity in its activity, i.e., a tendency to be preferentially recruited toward a specific direction. In general, hip muscle activity demonstrated significant directional specificity during hip efforts in contrast to knee and ankle muscle activity. Moreover, the angle of the mean vector calculated for each hip muscle across the sample remained relatively stable at both torque levels. These results indicate that directional patterns of muscle activation are unaffected by the level of torque produced. It is concluded that the method used in the present study is effective in characterizing directional patterns of muscle activation in a given population. Also, this method provides new perspectives for further quantification of muscle activation patterns disturbances in populations presenting various neuromuscular disorders.