RESUMEN
Bacterial infections have emerged as the leading causes of mortality and morbidity worldwide. Herein, we developed a dual-channel fluorescence "turn-on" sensor array, comprising six electrostatic complexes formed from one negatively charged poly(para-aryleneethynylene) (PPE) and six positively charged aggregation-induced emission (AIE) fluorophores. The 6-element array enabled the simultaneous identification of 20â bacteria (OD600=0.005) within 30s (99.0 % accuracy), demonstrating significant advantages over the array constituted by the 7 separate elements that constitute the complexes. Meanwhile, the array realized different mixing ratios and quantitative detection of prevalent bacteria associated with urinary tract infection (UTI). It also excelled in distinguishing six simulated bacteria samples in artificial urine. Remarkably, the limit of detection for E. coli and E. faecalis was notably low, at 0.000295 and 0.000329 (OD600), respectively. Finally, optimized by diverse machine learning algorithms, the designed array achieved 96.7 % accuracy in differentiating UTI clinical samples from healthy individuals using a random forest model, demonstrating the great potential for medical diagnostic applications.