Neuroimage 11, Number 5, 2000, Part 2 of 2 Parts 10 E h[@ ATTENTION Linking sensory and motor representations in working memory plan formation C. Richard Clark*, Kathryn Moores *, Alexander McFarlanet, Andrew Lewis*, Greg Brown+, Aina Puce& Gary Egan’l, James Brown+ *Cognitive Neuroscience Laboratory and School of Psychology, Flinders Universip, Adelaide, Australia; tDept of Psychiatry, Queen Elizabeth Hospital, Adelaide, Australia; $MRI Suite, Royal Adelaide Hospital, Adelaide, Australia $Brain SciencesInstitute, Swinbume University of Technology, Melbourne, Australia; ‘Howard Florey Institute, Melbourne, Australia An important task of the human central nervous system is to link sensory information to appropriate response. This is the defining characteristic of adaptive behaviour in humans. Such adaptability is presumed to be mediated by working memory systems that process and respond to detected stimuli according to experience, needs, context and intention. The prevent study examined brain processes associated with the establishment of intention in working memory The aim was to identify brain structures concerned with the establishment of sensori-motor contingency (execuhvr function) and to differentiate them from those concerned with the related. Table: Talairach coordinates of activations perceptual and behavioural represen- Stimulus updating Response updating tations (supramodal stores). Previous functional image work ha\ Region Intraparmtal sulcus ‘1 9, superior parietal gyms P0stcentral gyms Recentral gyms Precentral sulcus L‘ 9. “ I2 Middle occipital gyrus Superior frontal gyrus superi0r frontal sulcus Middle frontal gyms Middle frontal gyms Inferior frontal gyms ., 71 1‘ ,. Lateral orbital gyms Anterior cingulate sulcus Left Hem Right Hem Left Hem Right Hem established the significance of the pa- rietal and prefrontal cortex (PFC) in working memory function (see Clark -39 -35 45 36 -48 37 -38 -37 50 46 -35 41 et al, 2000). The inferior regions of the -24 -64 44 30 -61 49 -15 -71 40 parietal cortex are implicated in the 14 -67 52 -1 -62 54 18 -66 52 supramodal representation of percep- 46 -28 24 tual content whilst superior regions 40 -4 53 may be linked to the spatial localisa- -21 I 52 tion and imaginary mauipulation of -29 -4 52 such content. The ventral PFC in hu- -20 -8 55 34 5 47 -22 -83 1 mans appears to be required for the maintenance and comparison of mate- 10 16 56 rial in working memoty, whilst dorso- -11 -1 55 lateral PFC is recruited when addi- 26 7 58 37 38 24 39 34 29 tional monitoring or manipulation is -34 22 23 41 39 11 -35 13 10 34 25 10 required (Esposito et al, 1998). Whole brain FMRI was used to in- 37 27 10 48 11 22 vestigate brain function during the up- 37 49 5 dating of perceptual (stimulus) and be- havioural (response) representations 2 1748 -2 15 48 2 1552 relevant to a designated task goal. Ten normal subjects completed a visuover- bal task involving a finger response to infrequent targets. Target identity and prescribed response finger varied according to condition in a blocked design. In a fixed target condition, target identity was prescribed at the beginmng of the block. In a variable target condition, targets \cere defined as repeated words, necessitating the repeated updating of stimulus set for satisfactory task performance. In a fixed response condition, the response finger for targets was prescribed at the beginning of the block. In a variable response condition, enforced updating of response set was ensured by requrring the cycling the prescribed response finger with each attended word. Planned comparisons between conditions were undertaken to identify the brain regions involved in the updating of stimulus and response set for the same task. A pseudo-conjunction analysis was carried out to identify regions of common activation presumed to be involved in associating stimulus and response set. Analyses were carried out separately on each subject after motion correction using resel-corrected. voxel-based t-tests. Groups results were determined by identifying statistical consistency across subjects (half or more) in regional activation after alignment to a standard MR image using a 168df nonlinear algorithm (AIR 3). Peak activations (see Table) during the updating of stimulus set were obtained along the extent of the intraparietal sulcus UPS) bilaterally and extending into the postcentral, superior pa&al. supramarginal and angular pyri: the caudal aspects of the postcentral gyms bilaterally: bilaterally in the inferior (IFG) and middle (MFG) frontal gyri; the right superior frontal gyms: and lateral orhital gyms and medial superior frontal cortex. Response set updating produced activations along the IPS extending bilaterally into rhe wpramarginal gyms and oaudal postcentral gyms. and on the !eft mto the angular gyms: the right lateral and midline wperior parietal lobe t SPL): btlntcral motor and premotor regions: bilateral IFG; right MFG and medial superior frontal cortex. Areas ot act~~anon common to both type, ot updating mcluded the anterior aspects of the left IPS extendmg into the antenor inferior parietal lobe fIPL) and caudal po~tcmtrnl gyms. rowal aspecta of the righl SPL. aspects of the right middle and inferior frontal gyri and the midline premotor area. This study identifies a network concerned with establishing sensorimotor contingency in working memory plans m right IFG. left rostra1 IPS and suppIementary motor area. Another insight suggested by the study wa\ a division of function in the inferior parietal and inferior prefrontal regiona to storage of information relevant to perceptual and behavioural sets in working memory as well as to executive function. An unexpected result wa\ the lack of activation in left MFG during stimulus updating. References Clark, C. R., Egan, G. F., McFarlane. A. C., Morris. P.. Weber, D. lo.. Sonkkilla, C Soda, J, Updating working memory for words: a PET activation study. Human Brain Mapping. Volume 9(2), In press. Esposito MD, Aguitre GK. Zarahn E, Ballard D, Shin RK and Lease J (1998) Functional MRI studies of spatial and nonspatial working memory. Cognitive Brain Research 7(13): l-13. s77