The relationship between local and global processing and the processing of high and low spatial frequencies studied by event-related potentials and source modeling

Abstract
The processing of global and local elements and of low- and high-spatial frequencies are thought to be interrelated. Evidence for this stems from findings showing that brain localizations for global/local elements and for low/high spatial frequencies seem to overlap. The present study aimed to provide direct evidence that topographical differences between the processing of global and local visual elements can directly be explained by their spatial frequency content, and to study at which point in time this relation is present. This was done by studying the event-related potentials (ERPs) and source models elicited by unfiltered, low- or high-pass filtered hierarchical stimuli.
Results showed that performance for global and local targets was affected by removing low and high spatial frequencies, respectively. ERP data indicated that at 250 ms, there was an interaction between the processing of global/local targets and of spatial frequencies because at this time-point removal of low spatial frequencies decreased activity associated with the processing of global targets.
When localizing this effect, we found evidence implying that spatial frequency content indeed affected the brain region in which local/global targets were processed. Results implicated that the processing of global information depended on its low spatial frequency content, which was processed more laterally. Instead, processing of local information seemed to depend on its high spatial frequency content, which was processed more medially. Thereby, present results extend former results showing that global and local processing is dependent on spatial frequency and mapped retinotopically in the visual cortex.

Reference: 
M.A. Boeschoten, C. Kemner, J.L. Kenemans, H. van Engeland | 2015
In: Cognitive brain research : a section of Brain research devoted to the publication of cognitive and computational studies, ISSN 1872-6348 | 24 | 2 | juli | 228–236
http://www.sciencedirect.com/science/article/pii/S092664100500039X