AI Algorithm-Assisted SERS Detection of Levothyroxine Sodium in Urine by an Optoplasmonic Film

By harnessing the capture of target molecules at the metal surface via a highly confined electromagnetic field induced by surface plasmons, surface-enhanced Raman spectroscopy (SERS) has emerged as a sensitive technique for spectral analysis. However, the precise identification of biomarkers in urine using SERS remains a formidable challenge due to the low concentration of biomarkers and the complex nature of the urine composition. In this investigation, we developed an optoplasmonic film SERS substrate comprising silver nanowires (AgNWs) and silicon dioxide (SiO2) to tackle this issue. The AgNWs produce robust SERS signals, while SiO2 enhances laser focusing, facilitating the efficient detection of levothyroxine sodium (LS) in urine with an ultralow detection limit of 1 × 10–9 M. Leveraging artificial intelligence (AI) algorithms, this optoplasmonic substrate achieved precise LS identification in urine with a remarkable accuracy rate. This pioneering substrate holds substantial promise for the noninvasive, swift detection of target molecules in biofluids, heralding advancements in disease diagnostics and health monitoring.