Spatial clustering statistics of cortical maps.

Spatial autocorrelation analysis measured by Geary's C provides a global assessment of spatial organization versus random distribution. A stringent analysis of local similarity and parameter clustering was performed by determining average value differences of each polygon from its direct neighbors. Polygon similarity expresses the proportion of sites with significantly similar neighbors (range 0–1). Both global and local measures were validated by Monte-Carlo analysis. Bold: p<0.05; italic: not significant.