The first step of brain network analysis in fMRI is to detect regions of interest. The signals from these ROIs are then used to evaluate neural networks and quantify neuronal dynamics The two main methods to identify ROIs are based on brain atlas registration and clustering. In this work, both paradigms are combined by using a hierarchical two-level approach. In the first level, the anatomical parcellation is applied and then a clustering-based strategy within anatomical labels is performed. The proposed method is compared with a current approach based on spectral clustering.