AI-enabled, enzyme-integrated photonic crystal sensor for real-time noninvasive visual uric acid detection

Since uric acid (UA) serves as a critical biomarker for diagnosing metabolic disorders such as gout and hyperuricemia, developing noninvasive optical sensors for UA analysis provides significant advantages. However, conventional visual detection sensors encounter challenges, including insufficient long-term stability, low selectivity against interferents, and limited colorimetric resolution. Herein, we present a UA-responsive photonic crystal (PC) sensor that transduces enzymatic recognition into vivid structural color shifts. By integrating a crack-free polystyrene@graphene oxide (PS@GO) template with an enzyme-incorporated responsive hydrogel, the UA-PC sensor exhibits a rapid response (∼20 min), high sensitivity with a detection limit of 5 μM, and excellent anti-interference performance against common urinary constituents. The robust structural color enables intuitive visual readout, and an AI-enabled machine learning strategy is further introduced to capture images in color, enabling quantitative UA analysis via a smartphone-based platform. The system establishes a proof-of-concept visual sensing platform for daily health screening and home-based biochemical self-monitoring, addressing persistent limitations in stability, selectivity, and visual clarity.