A　large　number　of　studies　on　altered　functional　brain　network　tend　to　focus　only　on　the　alternation　in　topological　metric　of　functional　connectivity,　rather　than　on　the　details　of　graph　adjustment　that　cause　topological　metric　changes,　such　as　significant　adjusted　route　and　the　nodes　on　it.　In　this　paper,　we　first　used　the　brain　atlas　of　Dosenbach　to　generate　the　functional　brain　networks　of　the　21　participants　recruited　in　the　mental　arithmetic　experiment.　Then,　the　nodal　efficiency　of　each　brain　region　in　the　network　were　calculated　and　statistically　compared　between　mental　arithmetic　cognitive　states.　The　brain　regions　with　significant　alternation　in　nodal　efficiency　were　taken　as　seeds　for　searching　adjusted　routes.　The　brain　regions　that　have　significant　changes　in　network　efficiency　with　the　seed　nodes　were　considered　as　destined　nodes　of　the　relative　seed　nodes.　Finally,　the　details　of　two　adopted　indicators　on　altered　functional　brain　network　by　comparing　the　adjusted　route　between　the　two　endpoints　of　the　adjusted　route　were　given　and　used　as　clues　for　the　better　understanding　of　the　cognitive　pattern　of　mental　arithmetic.　In　this　paper,　the　average　number　of　adjusted　routes　contributed　by　brain　region　is　used　to　indicate　the　degree　of　contribution　of　the　brain　region　to　the　route　adjustment,　and　the　interaction　degree　within　specific　network　is　indicated　by　the　density　of　adjusted　routes.　The　results　show　that　both　indicators　of　fronto-parietal　network　is　significantly　higher　than　that　of　other　networks,　which　indicates　the　brain　regions　and　the　routes　within　fronto-parietal　network　are　the　most　active.　In　summary,　the　method　proposed　in　this　paper　provides　a　new　perspective　to　study　the　causes　of　functional　brain　network　alternation　in　mental　arithmetic.　However,　due　to　the　participants＇　variation　of　adjusted　routes　and　the　nodes　on　it,　a　better　understanding　of　these　functional　brain　network　alternation　for　individual　participant　with　the　proposed　method　needs　more　in-depth　research.