Analysis of sEMG during voluntary movement--Part II: Voluntary response index sensitivity

In this paper, a method for analyzing surface electromyographic (sEMG) data recorded from the lower-limb muscles of incomplete spinal-cord injured (iSCI) subjects is evaluated. sEMG was recorded bilaterally from quadriceps, adductor, hamstring, tibialis anterior, and triceps surae muscles during voluntary ankle dorsiflexion performed in the supine position as part of a comprehensive motor control assessment protocol. Analysis of the sEMG centered on two features, the magnitude of activation and the degree of similarity [similarity index (SI)] of the sEMG distribution to that of healthy subjects performing the same maneuver (n = 10). The analysis calculations resulted in response vectors (RV) that were compared to healthy-subject-derived prototype response vectors resulting in a voluntary response index (VRI). Incomplete SCI subjects (n = 9) were used to test the sensitivity of this analysis method. They were given supported-weight treadmill ambulation training, which is expected to improve or at least not cause a deterioration of voluntary motor control. The VRI provided evidence that the quantitative sEMG analysis method used was able to differentiate between healthy subjects and those with iSCI, characterize individual differences among iSCI subjects, and track motor control changes occurring over time.