Altered topological organization of high-level visual networks in Alzheimer's disease and mild cognitive impairment patients

Altered regional activation of high-level visual (HLV) cortices in patients with Alzheimer's disease (AD) or mild cognitive impairment (MCI) has been well documented in previous fMRI studies, which led us to investigate the underlying alteration of the HLV networks in the terms of intrinsic interaction and topological organization. First, the activation likelihood estimation, a coordinate-based meta-analysis approach, was used to define the cortical regions/nodes included in HLV networks of “what” and “where” visions. Secondly, the acquired HLV regions were used as seeds to calculate their interregional resting-state functional connectivities (RSFCs) based on the temporal correlation of rs-functional MRI (rs-fMRI) time series. Here, the rs-fMRI data of AD (n = 30), late MCI (n = 35), early MCI (n = 52) and matched healthy controls (n = 44) were obtained from the Alzheimer's Disease Neuroimaging Initiative dataset. Finally, based on the calculated pair-wise RSFCs, the “what” and “where” HLV networks were respectively constructed, and their topological properties were calculated and analyzed among groups using the graph theory method. The results demonstrated increased clustering coefficient combined with a prolonged characteristic path length of the “where” visual network in AD patients. No significant alternation of the “what” visual network was found among the groups. These results suggest that the abnormality of the HLV networks could be a late-stage outcome in AD and that the “where” visual network may be more susceptible to the AD-related neuropathological changes than the “what” visual network. In addition, the dysfunction of the “where” network is found to be characterized by a decreased integration combined with an increased local segregation.