Ultrarapid Kinetic Antimicrobial Susceptibility Testing from Blood Using Single-Cell Scattering Phenotypic Imaging

Bloodstream infections (BSIs) demand rapid antimicrobial susceptibility testing (AST) to guide early targeted therapy. However, current workflows rely on prolonged blood culture enrichment, delaying phenotypic guidance by 48-72 h. Here, we present a kinetic, single-cell nanoscale scattering phenotypic imaging platform that enables ultrarapid AST directly from complex blood samples. The large-volume imaging architecture provides mm³-scale observation with single-cell sensitivity. It allows simultaneous visualization and enumeration of tens to thousands of bacteria as individual nanoscale optical scatters without isolation, labeling, or microfluidic trapping. We introduce a growth-inhibition kinetic model that extracts effective growth parameters from dynamic single-cell population trajectories, enabling robust, mechanism-independent phenotypic classification across antibiotic classes. Using this approach, minimum inhibitory concentrations and categorical susceptibility are determined within 2-3 h for samples containing ≥10³ CFU mL^-1 and within <8 h for ultralow loads (∼2-10 CFU mL^-1) using only a brief preparatory growth step, without conventional blood culture. Applied to clinical blood-culture-positive BSI samples, the platform achieved 96% categorical agreement across 11 antibiotics in 100 tests, demonstrating reliability in clinically complex matrices. We further validated feasibility in whole blood by testing spiked samples across 2-10⁷ CFU mL^-1, confirming broad operating range, high detection sensitivity, and resilience to blood matrix interference. This reagent-minimal, imaging-only workflow enables clinically actionable AST turnaround within hours, offering a practical foundation for earlier precision therapy and improved antimicrobial stewardship in critical infection management.