Neurophysiological assessment of lower-limb voluntary control in incomplete spinal cord injury

Study design: Cross-sectional retrospective study of a neurophysiological method of voluntary motor control characterization. Objectives: This study was undertaken to validate the surface electromyography (sEMG)based voluntary response index (VRI) as an objective, quantitative, laboratory measure of spinal cord injury severity in terms of voluntary motor control disruption. Setting: VA Medical Centers in Houston and Dallas Texas, USA. Methods: A total of 67 subjects with incomplete spinal cord injury (iSCI), American Spinal Injury Association Impairment Scale (AIS)-C ( n = 32) and - D ( n = 35) were studied. sEMG recorded during a standardized protocol including eight lower-limb voluntary motor tasks was analyzed using the VRI method that relates multi-muscle activation patterns of SCI persons to those of healthy-subject prototypes ( n = 15). The VRI is composed of a measure of the amount of the sEMG activity ( magnitude) and the distribution of activity across muscle groups compared to that of healthy subjects for each motor task ( similarity index, SI). These resulting VRI components, normalized magnitude and SI, were compared to AIS clinical findings in this study. Receiver operating characteristic analysis was performed to determine the SI values best separating AIS-C and AIS-D subjects. Results: Magnitude and SI for AIS-C subjects had mean values of 0.27 +/- 0.32 and 0.65 +/- 0.21, respectively. Both parameters were significantly larger in the AIS-D subjects (0.78 +/- 0.43 and 0.93 +/- 0.06), respectively ( P<0.01). An SI value of 0.85 was found to separate AIS-C and AIS-D groups with a sensitivity of 0.89 and a specificity of 0.81. Further, the VRI of each leg strongly correlated with the respective AIS motor score (0.80, r<0.01). Conclusions: In the domains of voluntary motor control, the sEMG-based VRI demonstrated adequate face validity and sensitivity to injury severity as currently measured by the AIS.