Regional Homogeneity

Regional Homogeneity (ReHo) is an important aspect of neural network analysis used to measure the similarity of regional brain activity. It is part of the standard resting-state fMRI approach, a commonly used approach to measure brain activity and connectivity during rest. ReHo measure the degree to which signal intensity of MRI voxels located in the same anatomical region are similar. Along with other approaches, ReHo allows scientists to study the non-task related, or “background” activity in the brain, thus providing an insight to the functional organization of the brain at rest. The exact method to calculate ReHo involves measure the similarity of signal intensity between voxels located within the same cube. Each cube is defined by a cluster of the voxels of interest, usually of size 27 or 19. This is done by first calculating the Kendall's coefficient of concordance (KCC) of the voxels intensity, outputting a value between 0 and 1. Here, 0 is the lowest level of similarity among the nodes, while 1 indicates maximum regional similarity. ReHo is often used to locate resting state networks, such as the default-mode network, in which brain regions present remarkably similar resting-state activity. ReHo analysis is also used to diagnose brain disorders such as Alzheimer’s disease or depression, as research has found that differences in the ReHo results of healthy individuals compared to those with certain conditions such as autism and schizophrenia. In addition to its clinical applications, ReHo analysis is also used to identify brain regions which are most prone to plastic changes due to learning and development. Studies have suggested that certain brain regions that show greater fluctuation in ReHo from rest states to task states.