The corticotopic organization of the human basal forebrain as revealed by regionally selective functional connectivity profiles

Abstract

The cholinergic basal forebrain (CBF), comprising different groups of cortically projecting cholinergic neurons, plays a crucial role in higher cognitive processes and has been implicated in diverse neuropsychiatric disorders. A distinct corticotopic organization of CBF projections has been revealed in animal studies, but little is known about their organization in the human brain. We explored regional differences in functional connectivity (FC) profiles within the human CBF by applying a clustering approach to resting‐state functional magnetic resonance imaging (rs‐fMRI) data of healthy adult individuals (N = 85; 19–85 years). We further examined effects of age on FC of the identified CBF clusters and assessed the reproducibility of cluster‐specific FC profiles in independent data from healthy older individuals (N = 25; 65–89 years). Results showed that the human CBF is functionally organized into distinct anterior‐medial and posterior‐lateral subdivisions that largely follow anatomically defined boundaries of the medial septum/diagonal band and nucleus basalis Meynert. The anterior‐medial CBF subdivision was characterized by connectivity with the hippocampus and interconnected nodes of an extended medial cortical memory network, whereas the posterior‐lateral subdivision was specifically connected to anterior insula and dorsal anterior cingulate components of a salience/attention network. FC of both CBF subdivisions declined with increasing age, but the overall topography of subregion‐specific FC profiles was reproduced in independent rs‐fMRI data of healthy older individuals acquired in a typical clinical setting. Rs‐fMRI‐based assessments of subregion‐specific CBF function may complement established volumetric approaches for the in vivo study of CBF involvement in neuropsychiatric disorders.

Relationship between Basal Forebrain Resting-State Functional Connectivity and Brain Amyloid-β Deposition in Cognitively Intact Older Adults with Subjective Memory Complaints

“…We also investigated the role of sex and APOE genotype on the associations between basal forebrain RSFC and the global amyloid deposition. We disclosed a significant effect functional MRI functional connectivity profiles (29). First, we confirmed the presence of specific and differentiated ABF versus PBF functional networks.…”

“…and S.J.T.). The details of this independent methodologic study are described in the original article (29); we summarize here all the necessary information to understand the functional definition of the used basal forebrain seed regions. All voxels in a stereotactically defined basal forebrain region of interest (30) were clustered into functionally homogeneous partitions by applying a k-means clustering algorithm to the voxelwise brain-wide functional connectivity profiles as derived from resting-state functional MRI data (31).…”

“…First, we confirmed the presence of specific and differentiated ABF versus PBF functional networks. Reproducing the initial characterization of these distinct basal forebrain seeds (29), we found positive ABF RSFC with regions mainly belonging to anterior-posterior midline and medial temporal default mode network nodes, whereas positive PBF RSFC was mainly observed in the salience network, a system primarily consisting of bilateral anterior insular cortex, dorsal anterior cingulate cortex, and supplemental motor area (34,35). These results support the key role of the basal forebrain cholinergic system in executive processes, aligning previously described associations between basal forebrain atrophy and executive function deficits in prodromal stages of AD (36,37).…”

Functional parcellation of the hippocampus by semi-supervised clustering of resting state fMRI data

“…5). Simultaneously, the proposed method could generate parcellation results with significantly higher functional homogeneity than the structural brain parcellation method [1][2][3][4] , masked ICA based brain parcellation method 13,15,70 , and k-means clustering based brain parcellation method 37,71 , respectively (Fig. 5).…”

“…In our experiments, we applied the masked ICA based method to parcellate the hippocampus based on the estimated subject-specific ICs. Additionally, we compared the proposed semi-supervised method with k-means clustering based unsupervised method, which has been widely adopted in brain parcellation 37,71 . Particularly, elements a uv in the adjacency matrix A (i.e., functional similarity matrix) are defined in Eq.…”