Fast and reliable determination of Escherichia coli susceptibility to antibiotics: Infrared microscopy in tandem with machine learning algorithms

Abstract

Antimicrobial drugs have an important role in controlling bacterial infectious diseases. However, the increasing resistance of bacteria to antibiotics has become a global health care problem. Rapid determination of antimicrobial susceptibility of clinical isolates is often crucial for the optimal antimicrobial therapy. The conventional methods used in medical centers for susceptibility testing are time-consuming (>2 days). Two bacterial culture steps are needed, the first is used to grow the bacteria from urine on agar plates to determine the species of the bacteria (~24 hours). The second culture is used to determine the susceptibility by growing colonies from the first culture for another 24 hours. Here, the main goal is to examine the potential of infrared microscopy combined with multivariate analysis, to reduce the time it takes to identify Escherichia coli susceptibility to antibiotics and to determine the optimum choice of antibiotic to which the bacteria will respond. E coli colonies of the first culture from patients with urinary tract infections (UTI) were examined for the bacterial susceptibility using Fourier-transform infrared (FTIR). Our results show that it is possible to determine the optimum choice of antibiotic with better than 89% sensitivity, in the time span of few minutes, following the first culture.

Keywords: SVM; bacterial resistance to antibiotics; infrared spectroscopy; multivariate analysis.