Imaging studies on the functional organization and plasticity of human visual cortex

The visual system occupies approximately 25% of the human cerebral cortex. Large part of the visual cortex is organized into retinotopic maps, in which nearby cortical locations represent neighboring points in the visual field. Based on these retinotopic maps and on functional specialization for particular visual stimuli, such as movement or objects, human visual cortex is divided to more than 20 distinct visual areas. This Thesis contributes to our understanding on the functional organization and reorganization of the human visual cortex. In the first study, a method for retinotopic mapping based on multifocal functional magnetic resonance imaging (fMRI) responses was developed. Multifocal refers here to parallel stimulation of multiple regions in the visual field using temporally orthogonal stimulus sequences. Multifocal fMRI provides a straightforward analysis and interpretation of the retinotopic responses. The retinotopy in the primary visual cortex (V1), and in a subset of visual areas beyond V1, can be mapped with the multifocal fMRI. In a separate study, the nonlinear spatial summation of the multifocal responses in V1 was characterized. The interactions between adjacent visual field regions suppressed the multifocal responses, most likely due to the far-reaching spatial summation of V1 cells. The third study showed that training can reorganize visual cortices in an adult patient. The subject was a patient with a chronic visual field defect (homonymous hemianopia) due to a lesion in the visual cortex. We followed the intensive training of visual functions in his blind hemifield with recordings of evoked neuromagnetic responses. After successful training, fMRI measurements revealed an abnormal representation for the trained (right) visual field in the visual areas in the healthy (right) hemisphere, indicating large-scale reorganization of the visual processing. The fMRI tuning curves for spatial frequency were measured in the fourth study. The spatial frequency evoking the strongest fMRI response decreased with the eccentricity of the visual field. The differences in spatial frequency representation between visual areas support the view that these areas may process visual information at different spatial scales. Finally, the cortical sensitivity to phase relations between different spatial frequencies and the significance of congruent phase structure was revealed in multiple human visual areas. The results suggest that higher-level visual areas use the phase congruency information in the detection of natural broadband edges