Cortical areas involved in virtual movement of phantom limbs: comparison with normal subjects.

OBJECTIVE: To demonstrate that amputees performing "virtual" movements of their amputated limb activate cortical areas previously devoted to their missing limb, we studied amputees with functional magnetic resonance imaging (fMRI) and positron emission tomographic (PET) scans and compared the results with those of normal volunteers performing imaginary movements during fMRI acquisitions.
METHODS: Ten amputees (age range, 33-92 yr; average age, 49 yr; six men and four women; eight upper-limb and two lower-limb amputations) able to move their phantom limb at will were studied by fMRI (all patients) and PET scan (seven patients). The time between amputation and fMRI and PET studies ranged from 1 to 27 years (average, 13 yr). Patients were asked to perform virtual movements of the amputated limb and normal movements of the contralateral normal limb according to the functional images acquisition procedure. Movements of the stump were also used to differentiate stump cortical areas from virtual movement-activated areas. Ten right-handed volunteers, age- and sex-matched to the amputees, were also studied by fMRI. All volunteers were asked to perform four tasks during their fMRI study: imaginary movements of their right arm (1 task) and foot (1 task) and real movements of their left arm (1 task) and foot (1 task).
RESULTS: In amputees, virtual movements of the missing limbs produced contralateral primary sensorimotor cortex activation on both fMRI and PET scans. These activation areas, different from the stump activation areas, were similar in location to contralateral normal limb-activated areas. Quantitatively, in two amputees who claimed to be able to perform both slow and fast virtual movements, regional cerebral blood flow measured by PET scan in the precentral gyrus increased significantly during fast movements in comparison with slow virtual movements. In normal subjects, significant differences between real versus imaginary fMRI activations were found (for both foot and hand movements); imaginary right hand and foot tasks activated primarily the contralateral supplementary motor areas, with no significant activation detected in the contralateral precentral or postcentral gyri.
CONCLUSION: Primary sensorimotor cortical areas can be activated by phantom-limb movements and thus can be considered functional for several years or decades after amputation. In this study, we found that the location of the activation of these areas is comparable to that of activations produced by normal movements in control subjects or in amputees.