Imagined locomotion in the blind: an fMRI study

Functional magnetic resonance imaging (fMRI) in sighted individuals previously showed parahippocampal and fusiform activations during locomotor imagery, which were interpreted to reflect visuospatial navigation. Concurrent deactivations of multisensory vestibular and somatosensory cortical areas may reflect suppression of vestibular and somatosensory input, in order to prevent adverse interactions of sensory signals with the optimized automated locomotion pattern. In this fMRI study we compared blood oxygen level dependent (BOLD) activations and deactivations during the kinesthetic imagery of standing, walking, and running in seven congenitally totally blind subjects, seven sighted age-matched controls, and five subjects with age at onset of complete blindness > or =9 y or minimal residual vision. Imagined lying served as the rest condition. As opposed to their sighted controls, congenitally totally blind individuals activated multisensory vestibular areas in the posterior insula and superior temporal gyrus during imagined locomotion. Further, congenitally blind individuals did not show activations in parahippocampal and fusiform regions during locomotor tasks. In the intergroup comparisons, congenitally blind subjects exhibited higher BOLD activity levels than sighted subjects in multisensory vestibular (posterior insula and adjacent temporal sites), somatosensory (postcentral gyrus), and primary motor cortical areas, while sighted subjects showed higher activity levels in the parahippocampal and fusiform gyri. These findings indicate that blind subjects rely more on vestibular and somatosensory feedback for locomotion control than sighted subjects. This is accompanied by enhanced voluntary motor control and enhanced motor-kinesthetic processing. Thus, we provide neuroimaging evidence of distinct sensorimotor strategies in the blind for locomotor control.