Biofilm-Forming Ability of Infectious Organisms on Biomimetic Surfaces─An <i>In Vitro</i> and Machine-Learning Analysis

The current study explores the adhesion and biofilm-forming ability of different opportunistic pathogens including <i>Staphylococcus aureus</i>, <i>Escherichia coli</i>, <i>Lactobacillus</i> spp., <i>Streptococcus mutans</i>, and <i>Pseudomonas aeruginosa</i> on lotus leaf (LL) and peepal leaf (PL) inspired biomimetic hydrophobic surfaces. Surface topology that mimics the respective leaves was fabricated using polylactic acid by solvent casting. Water contact-angle measurements revealed varying degrees of material surface hydrophobicity with respect to the varying surface roughness. The biofilm formation was significantly influenced by the type of polymer surface (<i>p</i> < 0.005) and the hydrophobicity of the bacterial surface (<i>p</i> < 0.0001). Multilayer perceptron (MLP), a feed-forward neural network, gave the best results with 5-fold cross-validation and an accuracy of 85%. J48-base model predicted that organisms with a surface hydrophobicity of >57% had higher biofilm-forming ability than others. Similarly, polymers with low surface roughness (roughness < 0.46) had reduced biofilm formation. In conclusion, biomimetic hydrophobic surfaces reduce the biofilm formation on implants.