EP 34. Functional hierarchy within the neural network for optokinetic ‘look’ nystagmus

Abstract

Key nodes of neural networks for ocular motor control and visual motion processing have been localized using saccades, smooth pursuit, and optokinetic nystagmus (OKN). Within the context of an independent fMRI study using OKN, 9 bilateral network nodes were localized comprising cortical eye fields in frontal (FEF), supplementary motor (SEF), cingulate (CEF) and parietal cortex (PEF), visual motion centers MT+ and V6, the superior colliculus (SC), the lateral geniculate nucleus (LGN) and the globus pallidus (GP). Here, we examined the network’s functional hierarchy as present in the structural co-variation (SCoV) and resting-state (RS) fMRI, and the effect of RS condition (eyes open/closed) on its’ functional connectivity (FC). Two publicly available samples were analyzed consisting of the enhanced NKI sample with RS (TR 1.4s) and structural MR data (n = 124; age 46.7 ± 17.6; 40 male) and the “Beijing: eyes open eyes closed sample” measuring RS (TR 2s; n = 48; age 22.5 ± 2.2; 24 male). For the FC analysis, ICA-based denoising (FSL) was applied before spatial preprocessing (SPM) and band-pass filtering. Each bilateral ROI was represented by the first eigenvariate of the respective voxels’ time-series and partial correlation were computed using FSLNets. One group t-tests were computed over Fisher’s z transformed correlation coefficients. Each ROIs volume was approximated with voxel-based morphometry (VBM8) using non-linearly modulated gray matter density and partial correlations were computed for SCoV. Hierarchical cluster analysis was applied to determine sub-clustering within the OKN network. Edge-wise comparisons between RS conditions were performed using permutation testing and Bonferroni correction. Both FC and SCoV revealed two major subcluster. MT+ and V6 were similar to LGN and SC. The cortical eye fields clustered together with the GP. As effect of RS condition, with eyes closed the CEF switched to the visual subcluster. The edge-wise comparison revealed generally higher FC with eyes open and in particular a decrease of FC between MT+ and PEF, FEF and SEF as well as between V6 and SEF. Hierarchical clustering based on RS and structural data revealed a task-independent sub-division of the network for ocular-motor control and visual motion processing into two streams either involved in top-down (efferent voluntary) ocular-motor control (FEF, PEF, SEF, GP) and in more bottom-up visual target tracking (MT+, V6, LGN, SC) streams. This general network hierarchy was equally present in the RS with eyes open and eyes closed, with the CEF fulfilling a condition specific role in the network. The edge-wise comparison between RS conditions strengthens the evidence for a specific influence of MT+ on the ocular-motor control subcluster. These findings indicate a systematic influence of the resting condition not only on FC of the visual system, but on the state of the whole OKN network, while a general system hierarchy is omnipresent independent of RS condition.