Decoding urban street patterns in the human brain

We report on an experiment that used functional magnetic resonance imaging (fMRI) and multi-voxel pattern analysis (MVPA) in order to decode the neural correlates of urban street patterns in map reading. Participants were scanned while viewing real-world examples of organic, grid and mixed street network patterns. Three-quarters of data were used to train a classifier to distinguish the subject’s cognitive state while viewing the different type of street networks. The
remainder of the data was then used to test the classifier’s generalization performance. Results provide an important step in the understanding of pattern recognition as a fundamental component of geospatial thinking. They also provide evidence to what has been previously referred as the internalization or the urban street network (Hillier, 2003) and can have important implications for theories of urban form.