[Cognitive evoked potentials in face recognition in autism]

Face recognition requires a special type of processing that is different to that used to recognise objects. The right medial region of the fusiform gyrus displays a higher degree of activation when dealing with faces, which is why it is known as the fusiform face area, although the integrity of the right occipital face area may be needed to perform this task. Neuroimaging studies point not only to the existence of normal activation of the fusiform face area in response to faces but also of the occipital area. One modulation that supports the existence of a specific pathway is the face inversion effect, which is defined as the process in which greater disability is displayed in the recognition of inverted faces than for other mono-oriented objects. The P1 potential is the initial endogenic component of visual processing, acting as an index of the processing of features that show lower latency and greater amplitude when dealing with a face stimulus than when stimulated by objects. A later response, the negative component N170, also has lower latency in the case of faces than in inversion and objects. Its amplitude increases when the faces are inverted due to the generation of a greater activation in the region where it originates. A later negative potential can be associated to semantic identification in face recognition. The N170 waves in autistic individuals display greater latency with faces that is similar to the response to objects and, in turn, show no differences in the latencies with inverted faces. It remains to be shown whether they have an N400 response to faces.

Cognitive evoked potentials allow us to evaluate face recognition both in normal and autistic individuals by presenting responses to specific stimuli. Further studies will enable us to determine whether an N400 response exists in individuals with autism.