To construct brain-inspired AI models, cortical hierarchy is one aspect of brain anatomical information. Research is being conducted to estimate the cortical hierarchy in non-human primates, but brain-region names do not correspond across species. Therefore, in order to apply the knowledge of the cortical hierarchy in non-human primates to humans, it is necessary to match brain-region names between different brain atlases. Every brain atlas includes a catalog of brain regions, adjacency information among regions, and nomenclatural descriptors, and, despite individual variations, follows a broadly similar structural framework. Therefore, we aimed to associate brain-region names between different brain atlases by treating the structure of each brain atlas as a graph G with nodes V, an adjacency matrix \({{\boldsymbol{A}}}\) , and a node feature matrix \({{\boldsymbol{X}}}\) , and performing graph matching. Deep Graph Matching Consensus (DGMC) was used as the algorithm to solve the graph matching. In step 1, to estimate the human cortical hierarchy, we used DGMC to obtain a correspondence matrix \({{\boldsymbol{S}}}\) of nodes, which is necessary for matching brain-region names between different brain atlases. In step 2, we estimated the human cortical hierarchy converting the non-human primates cortical hierarchy with the correspondence matrix.