Haleeful Jud, Haya Aijaz, Nisar Ahmad Wani
Catheter-related urinary tract infections (CAUTIs) remain a critical healthcare-associated complication, motivating the development of antibacterial polymeric materials for catheter applications. In this study, zinc oxide (ZnO) nanoparticles prepared under two calcination conditions (designated as nominal <50 nm and nominal <100 nm size groups) were incorporated into polyurethane/polyethylene glycol (PU/PEG) films by solvent casting at loadings of 5, 10, 20, and 30 wt%. The materials were characterized using Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), water contact-angle measurements, and disk-diffusion assays against Escherichia coli and Staphylococcus aureus. The antibacterial results demonstrated concentration-dependent inhibition, reaching maximum inhibition zones of 17 mm against E. coli and 16 mm against S. aureus at 30 wt% nominal <50 nm ZnO. To rigorously evaluate structure-property-function relationships and prevent model overfitting on a compact dataset, an exploratory Random Forest regression model was integrated using Leave-One-Out Cross-Validation (LOOCV). The cross-validated model showed good fit within the studied dataset. Quantitative feature importance ranking confirmed that ZnO concentration is the main governing factor of antibacterial efficacy (90.6%), followed by surface contact angle (8.3%), whereas nominal particle size category (0.4%) and bacterial species (0.6%) contributed minimally to overall variance. The results demonstrate the potential of cross-validated machine learning for exploratory feature screening in biomaterial design and indicate that ZnO-incorporated PU/PEG films warrant further investigation for potential urinary catheter applications. Direct coating adhesion, mechanical durability, cytocompatibility, release kinetics, biofilm inhibition, and performance on actual catheter substrates require future evaluation.