Due to the complexity of clinical samples, rapid and reliable bacterial identification and antimicrobial susceptibility testing (AST) remain challenging. To address these challenges, we developed a colorimetric sensing platform for bacterial identification and AST in clinical samples based on bacterial metabolism-driven synthesis of gold nanoparticles (AuNPs) via hydrogen peroxide (H2O2) mediation. In this strategy, bacteria metabolic differences among bacterial species were converted into distinct colorimetric signals. Integrated with linear discriminant analysis (LDA), our developed sensing system enables automated and high-resolution profiling of bacterial species and strains. We achieved 100% classification accuracy for seven bacterial species in serum and urine and successfully differentiated nine Escherichia coli strains. For AST, the system correctly assessed antibiotic resistance profiles in six clinical isolates, reaching an overall accuracy of 97.62%. Unlike the conventional AuNP-based aggregation sensors, our approach is more user-friendly, robust against environmental variability, and directly reflects bacterial metabolic activities. By directly converting metabolic signatures to diagnostic outcomes, this "read-to-answer" sensor array offers a powerful and accessible solution for bacterial identification and AST, with broad applicability in clinical and field settings.
Lin et al. (2025) studied this question.