Cortical potentials during imagined movements in individuals with chronic spinal cord injuries.

A closed-loop model of motor control predicts that central deafferentation should disrupt cortical motor processes when imagining movements of paralyzed limbs. To test this prediction, event-related potentials (ERP) were recorded from the supplementary motor area and the primary sensorimotor area in individuals with paraplegia or quadriplegia as well as able-bodied controls during executed/attempted and imagined movements of the hand and foot. The cross-correlation of ERPs generated during hand movement and imagery was slightly negative for controls, moderate and positive for paraplegics, and high and positive for quadriplegics. The cross-correlation between foot movement and imagery was moderate for controls, moderate to high for paraplegic and high for quadriplegic groups. For hand tasks, ERPs were uncorrelated between controls and quadriplegics; for the foot tasks, the correlations were low between controls and both paraplegics and quadriplegics. Amplitudes and latencies of the ERP were also compared between movement and imagery and between the three injury groups. A biphasic waveform appears prior to and during movements in controls that is absent during imagery and when attempting/imagining movements of paralyzed limbs. Two hypotheses are proposed to explain the differences in cortical processing between movement and imagery and between injury groups. First, cortical motor processes are altered by the absence of kinesthetic feedback during attempted movement of a deafferented limb as well as during imagery. Second, inhibitory processes, present during imagined movements of an intact limb, may be weakened by a spinal cord injury (SCI) so that movement and imagery processes appear isomorphic. While the absence of kinesthetic feedback from deafferented limbs likely contributes to some variability in motor processing, the influence of an SCI on movement inhibition requires further testing.