Planning and Spatial Working Memory: a Positron Emission Tomography Study in Humans

TLDR

High‑level planning is thought to involve prefrontal cortex and striatum, as shown by neurosurgical and neurodegenerative studies. The study aimed to map the neural substrates of cognitive planning with PET imaging. Twelve healthy volunteers underwent PET scans while completing easy and difficult Tower of London tasks, a mnemonic variant, and a control condition to assess regional cerebral blood flow. Difficult planning elevated left dorsolateral frontal cortex and caudate activity, with caudate and thalamus specifically engaged when contrasting difficult versus simple planning, while mnemonic variants activated dorsolateral frontal and ventral frontopolar regions bilaterally, indicating distinct ventral frontal contributions to spatial working memory and supporting a frontostriatal network for high‑level planning.

Abstract

Abstract Previous work with both neurosurgical and neurodegenerative patient groups has suggested that high level planning is mediated by neural circuitry which includes both the prefrontal cortex and the striatum. In this study, the functional anatomy of cognitive planning was investigated further, using positron emission tomography (PET). Regional cerebral blood flow (rCBF) was measured in 12 normal volunteers while performing easy and difficult versions of (i) a modified Tower of London planning task; (ii) a mnemonic variant of this task that required short‐term retention and reproduction of problem solutions; and (iii) a control condition that involved identical visual stimuli and motor responses. Significant increases in rCBF were observed in the left hemisphere, in both the mid‐dorsolateral frontal cortex and in the head of the caudate nucleus, when the difficult planning task was compared with the control condition. Moreover, subtraction of a simple planning condition from the more difficult one revealed focal increases in rCBF in the caudate nucleus and the thalamus only. During both mnemonic variants of the planning task, changes were also observed in the mid‐dorsolateral frontal cortex and in more ventral frontopolar regions, bilaterally. When compared directly, the planning and memory conditions differed in terms of these ventral activation foci, but not in the pattern of activation observed in the mid‐dorsolateral frontal cortex. These findings further implicate frontostriatal circuitry in high‐level planning and provide evidence for functionally distinct contributions from ventral and dorsolateral frontal regions to spatial working memory.