Photothermal Lateral Flow Assay with Inertial Microfluidic Enrichment for Early Detection of E. coli in Urinary Tract Infections

Abstract

Escherichia coli is the leading cause of urinary tract infections (UTIs), yet rapid and specific diagnosis at the point of care remains challenging. Conventional diagnostic methods, including urine culturing and microscopy, are time-intensive, require trained personnel, and often lack specificity for E. coli, limiting their effectiveness in resource-limited or time-critical settings. While lateral flow assays (LFAs) offer simplicity and portability, their clinical utility is constrained by a high visual limit of detection (LOD). This thesis presents an integrated diagnostic platform that combines passive microfluidic bacterial enrichment with photothermal detection to significantly improve LFA sensitivity for UTI diagnostics. Bacterial preconcentration was achieved using viscoelastic forces generated by flowing polyethylene oxide (PEO) solutions through custom-fabricated PDMS microchannels with straight, symmetric zigzag, and asymmetric zigzag geometries. Optimal enrichment was obtained in a straight 25 µm × 25 µm channel operated at 2 µL/min with 1000 ppm PEO, resulting in approximately an order-of-magnitude increase in bacterial concentration through sample volume reduction while retaining target cells. In parallel, a photothermal detection approach based on lock-in thermography was developed to detect temperature modulations arising from nanoparticle–target interactions on LFA test and control lines. This method enabled quantitative signal extraction beyond visual inspection and provided an additional two-order-of-magnitude sensitivity enhancement. Together, the integrated enrichment and photothermal detection platform achieved an overall ~1000-fold sensitivity improvement compared to standard visual LFAs. These results demonstrate the strong potential of combining passive microfluidic preconcentration with photothermal readout for rapid, sensitive, and reliable point-of-care UTI diagnostics. Identified limitations include manual operation, reliance on desktop-based processing, and pressure constraints of PDMS microchannels. Future work should focus on validation with clinical urine samples, platform automation with onboard fluidic and signal processing, evaluation of rigid microchannels for higher-pressure operation, and extension to additional bacterial pathogens and diagnostic applications.