Author

Benjamin Hu

Date of Award

December 2024

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Biomedical Engineering

First Advisor

Hossein Kashani Zadeh

Abstract

Biofilm are aggregates of microbial cells encased under an extracellular polymeric substance (EPS) and resist sanitizers or antibiotics. Because of their persistence on various surfaces such as medical devices or industrial piping, biofilm formation is associated with approximately 65% of all microbial infections. Thus, their detection and deactivation are of great importance to public health. Many detection methods such as swabbing require sampling by rubbing , which can be inefficient and challenging. This study proposes the use of handheld multispectral fluorescence imaging systems to detect and deactivate biofilm in situ. We utilized the Contamination and Sanitization Inspection (CSI) and -Disinfection (CSI-D+), both of which feature various cameras and UV-emitting LEDs to image and disinfect contaminations. Dual species Listeria monocytogenes/Pseudomonas aeruginosa or Shiga toxin-producing Escherichia coli (STEC)/P. aeruginosa biofilms were grown on polyvinyl chloride (PVC) and stainless steel (SS) coupons using an industrial bioreactor. We utilized 275 nm, 365 nm, or 405 nm illumination with different combinations of cameras to image biofilm formation. Further, we exposed biofilms to UVC (275 nm) irradiation (3.440 mW/cm2 power at a 1-foot distance) for 30, 60, or 130 sec, and the remaining bacteria were quantified with colony forming units (CFU). The combinations of 365 nm excitation with the camera with fluorescence triple bandpass filter 464/542/639 nm and 405 nm excitation with the camera with dual bandpass filter 535/700 nm of the CSI resulted in the best contrast between biofilm and the background surface. Deactivation efficacy of UVC was superior on SS compared to PVC which increased with exposure time. Up to 3.4 log CFU/cm2 removal of P. aeruginosa was observed from SS compared to ~1.4 log CFU/cm2 deactivation of STEC and L. monocytogenes, from PVC. Additionally, we utilized the color data from each image to train LDA model in discerning between biofilm and media with 92.5% accuracy. This technology can significantly improve biofilm detection and deactivation processes in various industries.

Available for download on Saturday, January 17, 2026

Share

COinS