TABLE 4.
Top 20 nodal graph theory features with maximum discriminative ability selected by group-LASSO.
| Nodal graph measures | AAL number | Corresponding brain regions | Anatomical classification |
| Nodal efficiency | 30 | Insula_R | Subcortical |
| Nodal clustering coefficient | 70 | Paracentral_Lobule_R | Parietal |
| Nodal shortest path | 11 | Frontal_Inf_Oper_L | Prefrontal |
| Nodal shortest path | 13 | Frontal_Inf_Tri_L | Prefrontal |
| Nodal shortest path | 29 | Insula_L | Subcortical |
| Betweenness centrality | 70 | Paracentral_Lobule_R | Parietal |
| Nodal efficiency | 29 | Insula_L | Subcortical |
| Nodal shortest path | 30 | Insula_R | Subcortical |
| Betweenness centrality | 38 | Hippocampus_R | Temporal |
| Nodal clustering coefficient | 36 | Cingulum_Post_R | Parietal |
| Nodal shortest path | 61 | Parietal_Inf_L | Parietal |
| Nodal local efficiency | 70 | Paracentral_Lobule_R | Parietal |
| Betweenness centrality | 28 | Rectus_R | Prefrontal |
| Degree centrality | 27 | Rectus_L | Prefrontal |
| Nodal shortest path | 14 | Frontal_Inf_Tri_R | Prefrontal |
| Nodal shortest path | 8 | Frontal_Mid_R | Prefrontal |
| Nodal shortest path | 9 | Frontal_Mid_Orb_L | Prefrontal |
| Nodal shortest path | 17 | Rolandic_Oper_L | Frontal |
| Betweenness centrality | 32 | Cingulum_Ant_R | Prefrontal |
| Betweenness centrality | 30 | Insula_R | Subcortical |
AAL, automated anatomical labeling atlas; LASSO, least absolute shrinkage and selection operator.