Authors:

BOR, D. & OWEN, A.M.

Reference:

Year of publication:

2005

CBU number:

6155

Abstract:

Most neuroimaging studies investigating the lateral prefrontal cortex (LPFC) report an association of LPFC activity with task demand. However, two recent studies provide important counter-examples. Using fMRI, in both the spatial and verbal domains, sequences of stimuli were presented for subjects to retain during a short interval and then retrieve. Stimuli were either randomly arranged or structured. While subjects performed the structured task better by strategically reorganising or "chunking" them into more efficient forms, LPFC activity was greater for the easier structure, compared with the harder random sequences. However, these two experiments involved structures that could be recoded either based on their mnemonic or mathematical content. The current study sought to investigate the differential level of prefrontal activity for purely mnemonic or purely mathematical chunking. During fMRI scanning, subjects carried out 3 conditions. Each condition shared the common format of presenting subjects visually with 4x2-digit numbers in sequence, requiring subjects to retain the 4 items over a short delay, and then make a verbal response demonstrating accurate retrieval of the sequence. One of these three conditions presented an entirely random sequence; a second involved sequences with a purely mathematical pattern, while a third incorporated non-mathematical sequences over-learnt during prior behavioural training. Both the mathematical and memory-based conditions activated LPFC, when compared with the random condition. However, mathematical trials activated the LPFC significantly more than memory-based trials. These results suggest that while the LPFC plays a general role in chunking strategies, it has a preferential involvement for mathematical chunking processes.