Accurate and rapid antibiotic susceptibility testing using a machine learning-assisted nanomotion technology platform

Antimicrobial resistance (AMR) is a major public health threat, reducing treatment options. A rapid growth-independent phenotypic AST using nanomotion technology measures bacterial vibrations. Machine learning techniques analyze a large dataset, achieving 90.5-100% accuracy. Independent testing on 223 strains predicts susceptibility and resistance with accuracies between 89.5% and 98.9%, demonstrating potential for future bacterial phenotype delineation.