The　presence　of　dysgraphia　in　patients　with　Parkinson＇s　disease　(PD)　is　frequently　observed　as　an　early　motor　symptom,　serving　as　a　vital　biomarker　for　early　diagnosis　and　monitoring　of　disease　progression.　However,　accurately　identifying　and　quantifying　dysgraphia　remains　a　significant　obstacle　for　individuals　affected　by　PD.　This　study　introduces　a　novel　quantification　approach　utilizing　a　16　×　16　pressure-sensitive　array　that　exhibits　exceptional　sensitivity　(0.72　N−1　at　1.02–4.14　N),　rapid　response　time　(10　ms),　sustained　signal　stability　(resistance　increased　by　2　%　after　1000　cycles),　and　reliable　electromechanical　synchronization.　Through　the　implementation　of　an　isolation　layer,　the　proposed　sensing　array　effectively　addresses　the　issue　of　signal　crosstalk　between　sensing　pixels　in　conventional　arrays,　thereby　enabling　precise　detection　of　subtle　irregularities　during　the　writing　process.　The　utilization　of　Resnet-18　CNN　in　conjunction　with　pressure-sensitive　array　yields　a　high　degree　of　precision　in　classification　for　Parkinson＇s　dysgraphia.　This　approach　offers　a　valuable　tool　for　detecting　dysgraphia　in　PD　patients.　©　2025