Sound-Encoded Faces Activate the Left Fusiform Face Area in the Early Blind

Abstract

Face perception in humans and nonhuman primates is accomplished by a patchwork of specialized cortical regions. How these regions develop has remained controversial, some studies claiming that seeing faces is necessary for face-domain formation, while others contend that visual experience is not necessary for their development. In sighted individuals, facial information is primarily conveyed via the visual modality. Blind individuals, on the other hand, can still recognize shapes using auditory and tactile cues. Here we examined whether face-selective regions like the fusiform face area (FFA) show selectivity for faces in individuals with early blindness. In particular, we asked if such individuals can learn to distinguish faces from other shapes by using a sensory substitution device (SSD) presenting faces as sound-encoded stimuli. Early blind individuals and sighted controls were trained to discriminate sound-encoded faces, houses, and abstract geometrical shapes presented via an SSD. Combining functional magnetic resonance imaging (fMRI) with an N-back task, we found new evidence for preferential activation of the left FFA by sound-encoded faces in early blind adults. The results demonstrate that face-selective cortical regions can be activated by substituting vision with audition. This implies that the development of the FFA does not necessarily depend on experience with visual faces per se, but may instead depend on sensory experience with facial configurations.