Cortical reorganization after spinal cord injury: An fMRI study
Abstract
To assess the degree of cortical reorganization in human SCI, somatotopic organization of the sensorimotor cortices was quantified during ankle sensorimotor tasks using fMRI. The purpose of this study was to identify the changes in the localization and extent of cortical activation in people with chronic SCI compared to neurologically intact controls. We hypothesized that the cortical somatotopic map would be largely intact in SCI due to residual spinal pathways, but that the area of cortical activation would be increased to partially compensate for the decrease in available spinal pathways. Cortical sensorimotor maps were measured in five neurologically intact (NI) subjects and five SCI subjects using fMRI during active, passive, and assisted movements of the ankle in the sagittal plane. The ankle was oscillated through a 10° range at 1Hz using a custom, nonmetallic device that secured the foot and lower leg minimizing out of plane motion. Cortical activation was seen in all subjects and generally localized to the expected areas of the sensory and motor cortices. Targeted activation of the ankle region was seen in primary somatosensory and motor cortices in all NI subjects. Although SCI subjects produced a localized area of activity, approximately corresponding to the location of the ankle, more fMRI signal was observed outside of the leg region of the cortex in SCI, compared to NI subjects. There was a higher activation mass during an active task than a passive task in both primary motor and somatosensory cortices in NI subjects. Also, there was a higher overall fMRI signal in the motor cortex in the SCI subjects, particularly for the active and assist tasks. In the lower leg region of the primary motor cortex, the SCI population showed consistently higher activation mass than the NI population for all three movement types; an average of 2.3 times higher (p=0.008). Conversely, in the lower leg primary somatosensory cortex, no statistically significant difference was found between the two populations (p<0.05, 2 factor ANOVA). The locations of the centers of activation were slightly different for the SCI subjects compared to NJ controls. The activation center had a greater spread for the SCI population and tended to be more lateral than the locations for the NI population.
This paper has been withdrawn.