Brain Topological Reorganization Associated with Visual Neglect After Stroke

BACKGROUND/PURPOSE: To identify brain hubs that are behaviorally relevant for neglect after stroke as well as to characterize their functional architecture of communication. METHODS: Twenty acute right hemisphere damaged patients underwent neuropsychological and resting-state functional magnetic resonance imaging sessions. Spatial neglect was assessed by means of the Center of Cancellation on the Bells Cancellation Test. For each patient, resting-state functional connectivity matrices were derived by adopting a brain parcellation scheme consisting of 153 nodes. For every node, we extracted its betweenness centrality (BC) defined as the portion of all shortest paths in the connectome involving such node. Then, neglect hubs were identified as those regions showing a high correlation between their BC and neglect scores. RESULTS: A first set of neglect hubs was identified in multiple systems including dorsal attention and ventral attention, default mode, and frontoparietal executive–control networks within the damaged hemisphere as well as in the posterior and anterior cingulate cortex. Such cortical regions exhibited a loss of BC and increased (i.e., less efficient) weighted shortest path length (WSPL) related to severe neglect. Conversely, a second group of neglect hubs found in visual and motor networks, in the undamaged hemisphere, exhibited a pathological increase of BC and reduction of WSPL associated with severe neglect. CONCLUSION: The topological reorganization of the brain in neglect patients might reflect a maladaptive shift in processing spatial information from higher level associative-control systems to lower level visual and sensory–motor processing areas after a right hemisphere lesion. IMPACT STATEMENT: In this study, by employing the theoretic graph measures of betweenness centrality, we identified a set of neglect hubs, that is a cohort of cortical hubs that are behaviorally relevant for neglect after right hemisphere stroke. Although regions in higher level associative-control systems decreased their centrality and efficacy in communication, lower level visual and sensory–motor processing areas exhibited a dysfunctional increase in centrality. These findings have clinical implications as neglect hubs may be targeted for noninvasive brain stimulation protocols to restore their centrality and in turn to ameliorate the spatial deficit.