Label-free phenotypic antimicrobial susceptibility testing on microfluidic platforms: a review of advances and translation.

The escalating global threat of antimicrobial resistance (AMR) necessitates a paradigm shift towards rapid and scalable diagnostic technologies. Conventional antimicrobial susceptibility testing (AST) methods, while reliable, are hindered by prolonged turnaround times (often 16-24 h) and complex workflows, limiting their utility at the point-of-care (POC). Label-free phenotypic AST methods integrated within microfluidic platforms present a compelling alternative, enabling real-time, high-throughput monitoring of bacterial responses without the need for costly or time-consuming fluorescent labels and reagents. This review provides a comprehensive analysis and comparative evaluation of four primary label-free detection modalities: electrical impedance sensing, light scattering, surface-enhanced Raman spectroscopy (SERS), and machine vision powered by artificial intelligence (AI). We critically examine the principles of each technique, its integration with microfluidics, and its performance, highlighting studies that have demonstrated AST results in critically short timeframes (< 2-4 h) with high accuracy. Furthermore, we uniquely synthesize progress in commercial translation, discuss persistent challenges such as standardization and sample preparation, and provide a strengths, weaknesses, opportunities, and threats (SWOT) analysis to chart the future trajectory of the field. By offering a critical roadmap that bridges fundamental innovation with clinical application, this review demonstrates the potential of label-free microfluidic systems to deliver reagent-free, rapid, and scalable diagnostic solutions in the fight against AMR.