AI-Assisted Viable Digital Array for Rapid Quantification of Bacteria in Pediatric Urinary Tract Infections

Abstract

Urinary tract infections (UTIs) are among the most common bacterial infections in neonates and infants, and delayed diagnosis can lead to severe complications. Rapid, quantitative detection of viable bacteria at the point of care remains a critical unmet need. Here, we present the Viable Digital AI-Integration SlipChip (ViDAI-SlipChip), a portable platform that integrates nanoliter-droplet digital culture, smartphone-based fluorescence imaging, and an artificial intelligence classifier (Hexa-MLP) for rapid absolute quantification of viable bacteria in urine samples. Using Escherichia coli as a representative model pathogen, the platform enables direct analysis of clinical urine samples without culture enrichment or nucleic acid extraction. The device partitions samples into 5-nL droplets containing a fluorogenic β-D-glucuronidase substrate; viable cells hydrolyze the substrate to produce localized fluorescence, enabling digital enumeration without preculture or nucleic acid extraction. Clinical validation with 120 pediatric urine samples demonstrated 94.4% sensitivity, 100% specificity, and excellent agreement with standard culture (overall accuracy 95.83%), achieving detection within 4 h. The Hexa-MLP classifier robustly interprets complex fluorescence patterns, overcoming limitations of conventional thresholding. By combining microfluidic confinement, accessible imaging, and AI-enhanced analysis, the ViDAI-SlipChip offers a rapid, accurate, and low-cost solution for point-of-care UTI screening, particularly in resource-limited settings.