Mapping the network for planning: a correlational PET activation study with the Tower of London task.

We used the Tower of London task (TOL) and H(2)(15)O-PET to map the network of brain structures involved in planning. Six healthy right-handed subjects had 12 measurements of relative regional cerebral blood flow (rrCBF) during six conditions, each performed twice. There was one rest condition, and five sets of TOL problems at different complexity levels, performed on a touch-sensitive computer monitor with the right arm. Complexity was defined as the number of moves required to solve each problem. Activation was analysed in two ways: a category analysis comparing levels of rrCBF during rest and task was done to identify all structures involved in performance of the TOL; and a correlation analysis was carried out to delineate a subset of structures where the levels of rrCBF correlated with task complexity. Activated brain areas in which rrCBF increases did not correlate with complexity could be grouped into: (i) regions belonging to the dorsal stream of visual input processing, namely visual cortical areas 17, 18 and 19, and posterior parietal cortical areas 7 and 40; and (ii) regions involved in the execution and sequencing of arm movements (right cerebellum, left primary motor cortex and supplementary motor area). Brain regions where levels of rrCBF correlated with task complexity included lateral premotor cortex (area 6), rostral anterior cingulate cortex (areas 32 and 24), dorsolateral prefrontal cortex (areas 9 and 46) bilaterally, and right dorsal caudate nucleus. We propose that dorsolateral prefrontal, lateral premotor, anterior cingulate and caudate areas form a network for the planning of movement that interacts with brain areas primarily involved in visual processing and movement execution.