AI Algorithm-Assisted\nSERS Detection of Levothyroxine\nSodium in Urine by an Optoplasmonic Film

By\nharnessing the capture of target molecules at the\nmetal surface\nvia a highly confined electromagnetic field induced by surface plasmons,\nsurface-enhanced Raman spectroscopy (SERS) has emerged as a sensitive\ntechnique for spectral analysis. However, the precise identification\nof biomarkers in urine using SERS remains a formidable challenge due\nto the low concentration of biomarkers and the complex nature of the\nurine composition. In this investigation, we developed an optoplasmonic\nfilm SERS substrate comprising silver nanowires (AgNWs) and silicon\ndioxide (SiO<sub>2</sub>) to tackle this issue. The AgNWs produce\nrobust SERS signals, while SiO<sub>2</sub> enhances laser focusing,\nfacilitating the efficient detection of levothyroxine sodium (LS)\nin urine with an ultralow detection limit of 1 × 10<sup>–9</sup> M. Leveraging artificial intelligence (AI) algorithms, this optoplasmonic\nsubstrate achieved precise LS identification in urine with a remarkable\naccuracy rate. This pioneering substrate holds substantial promise\nfor the noninvasive, swift detection of target molecules in biofluids,\nheralding advancements in disease diagnostics and health monitoring.