Schizophrenia is associated with well-documented deficits in high-order cognitive processes such as attention and executive functioning. The integrity of sensory-level processing, however, has been evaluated only to a limited degree. Our study evaluated the ability of patients with schizophrenia to recognize complete objects based on fragmentary information, a process termed perceptual closure. Perceptual closure processes are indexed by closure negativity (Ncl), a recently defined event-related potential(ERP) component that is generated within the visual association cortex. This study assessed the neural integrity of perceptual closure processes in schizophrenia by examining Ncl generation. Generation of the preceding positive(P1) and negative (N1) ERP components was also examined.
We evaluated 16 patients with chronic schizophrenia and 16 healthy comparison subjects. Successively less fragmented images were presented during high-density ERP recording, which permitted the monitoring of brain activity during perceptual closure processes prior to object recognition. Analyses were performed at parieto-occipital and occipitotemporal sites consistent with dorsal and ventral stream generators of P1, N1, and Ncl.
Patients with schizophrenia showed significant impairment in the ability to recognize fragmented objects, along with impaired generation of Ncl. The amplitude of visual P1 was significantly reduced, particularly over dorsal stream sites. In contrast, the generation of visual N1 was intact.
Patients with schizophrenia are profoundly impaired in perceptual closure as indicated by both impaired performance and impaired Ncl generation. The selective impairment in dorsal stream P1 is consistent with prior reports of impaired magnocellular processing in schizophrenia. By contrast, intact ventral N1 generation suggests that the initial stages of ventral stream processing are relatively preserved and that impaired magnocellular dorsal stream functioning in schizophrenia may lead to secondary dysregulation of ventral stream object recognition processing.