Recent　findings　in　the　field　of　neuroimaging　have　greatly　enriched　our　understanding　of　the　process　of　human　inductive　reasoning.　However,　as　an　extremely　complex　cognitive　activity,　inductive　reasoning　involves　multiple　cognitive　functions　such　as　attention,　memory,　executive　function,　symbol　processing,　and　fluid　intelligence,　and　requires　the　collaborative　participation　of　multiple　brain　regions.　Therefore,　achieving　a　unified　and　transparent　understanding　of　human　inductive　reasoning　from　different　datasets　presents　formidable　challenges.　This　paper　proposes　a　multi-loop　based　integration　analysis　framework　of　multiple　source　brain　data.　By　employing　representational　similarity　analysis　and　multi-task　functional　neuroimaging　analyses,　this　framework　combines　internal　evidence,　i.e.,　multi-source　functional　neuroimaging　data,　and　external　evidence,　i.e.,　knowledge　from　published　neuroimaging　articles,　to　reveal　different　neural　mechanisms　of　inductive　reasoning　in　DLPFC　and　SPL.　Experimental　results　reveal　the　intricate　uncertainty　surrounding　brain　localization　in　human　reasoning.　©　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.　2025.