EP 61. Coactivation-based parcellation of the posterior medial frontal cortex. Issue 9 (September 2016)

• Record Type: Journal Article
• Authors: Müller, V.; Cieslik, E.; Palomero-Gallagher, N.; Laird, A.; Fox, P.; Eickhoff, S.
• Is Part Of: Clinical neurophysiology. Volume 127:Issue 9(2016:Sep.)
• Page Start: e264
• Publication Date: 2016-09
• Subjects: Neurophysiology -- Periodicals; Electroencephalography -- Periodicals; Electromyography -- Periodicals; Neurology -- Periodicals; 612.8
• DOI: 10.1016/j.clinph.2016.05.113
• Languages: English
• ISSNs: 1388-2457

Abstract

Abstract : Introduction: The posterior medial frontal cortex (pMFC) is part of the so-called salience network (Seeley, 2007 ) and involvement of this region has been reported in a number of different tasks. Although usually referred to as dorsal anterior cingulate cortex, pMFC activation found in most studies is not restricted to the (middle) cingulate cortex, but also extends onto the dorsally abutting superior frontal gyrus, which in turn encompasses the pre-supplementary motor area and parts of the supplementary motor area. Using a connectivity-based parcellation approach, we aimed to investigate whether the pMFC can be subdivided with regard to its functional connectivity (FC) profiles and function. Methods: A volume of interest (VOI) corresponding to the pMFC was derived by the conjunction of three previous meta-analyses of attention, working memory and action inhibition (Müller, 2015 ). The whole-brain co-activation patterns for each voxel within the VOI were computed. Based on those co-activation patterns, all voxels of the VOI were grouped into distinct clusters using k-means clustering, and the most stable clustering solution identified. Meta-analytic and resting state FC approaches were then combined to examine the derived clusters' brain-wide connectivity. Finally, quantitative functional characterization was performed on each cluster in order to identify differentiable functions of the sub-clusters. Results: Results supported the segregation of the pMFC-VOI into five clusters: dorsal-anterior (C1), dorsal-posterior (C2), dorsal-middle (C3), ventral-anterior (C4) and ventral-posterior (C5). Importantly, results demonstrate a clustering into cingulate (C4, C5) and non-cingulate clusters (C1, C2, C3). Posterior cingulate cluster C5 was found to be mainly involved in pain processing and strongly connected to regions of the "pain matrix", while cingulate C4 more strongly relates to behavioral inhibition. Furthermore, within the non-cingulate pMFC, our results reveal an anterior higher order cognitive (C1) and a posterior action (C2) related cluster. Additionally, we found a central cluster (C3) revealing an intermediate connectivity profile and associated with working memory and cognitive flexibility tasks. Conclusion: Thus, our results show that, based on their connectivity patterns, cingulate and non-cingulate regions of the pMFC can be distinguished and further demonstrate a motor to cognitive clustering within non-cingulate pMFC and two cingulate clusters differently involved in pain processing and behavioral inhibition.