Non-invasive, in-situ and culture-free detection of bacterial conjunctivitis pathogens in trace tears using a capillary-integrated SERS sensor and portable Raman spectrometer

Bacterial conjunctivitis threatens visual function and requires rapid, ultrasensitive diagnostic approaches. Conventional conjunctival swab-based diagnosis is invasive and relies on time-consuming microbial culture. Herein, a capillary-integrated surface-enhanced Raman spectroscopy (SERS) platform was developed for non-invasive tear sampling and rapid on-site detection using a handheld Raman spectrometer. The capillary-integrated SERS sensor achieves an enhancement factor of 10 10 , with a limit of detection of 10 3 CFU/ mL for Staphylococcus epidermidis . Characteristic spectral features of infected tear samples were extracted to establish biochemical barcodes, which were matched with standard bacterial fingerprint barcodes for rapid pathogen typing of ocular infections. Combined with support vector machine (SVM) analysis, the system achieves 89% accuracy in clinical pathogen identification. This platform requires only 5 μL of tears and completes the assay within 5 min. Its enclosed capillary architecture also facilitates sample storage and data management, offering a cost-effective, non-invasive strategy for point-of-care diagnosis of ocular infectious diseases.