Flexible　organic　synaptic　transistors　(FOSTs)　are　crucial　for　neuromorphic　computing　due　to　their　flexibility　and　biocompatibility,　yet　their　mechanical　stability　under　strain　is　underexplored.　This　study　enhances　FOST　resilience　by　optimizing　the　neutral-axis　alignment　through　layer　thickness　adjustments　and　incorporation　of　a　polyimide　layer,　aligning　the　axis　closer　to　the　heterojunction　interface.　This　strategy　significantly　reduces　strain-induced　defects,　minimizing　excitatory　postsynaptic　current　(EPSC)　degradation　from　21.19%　to　13.34%　after　100　bending　cycles.　Optimized　FOSTs　demonstrate　a　remarkable　pattern　recognition　accuracy　of　90.4%　after　bending,　significantly　outperforming　the　76.8%　achieved　by　standard　devices.　These　findings　present　a　straightforward　and　effective　approach　to　improve　the　mechanical　stability　and　synaptic　performance　of　FOSTs,　advancing　the　development　of　durable　bio-inspired　computing　systems.