Rapid Strain Differentiation of E. coli-inoculated Urine Using Olfactory-based Smart Sensors

Urinary tract infections (UTIs) have a large effect on the public healthcare system. Millions are affected each year, and are likely to have the infection recur with increased resistance to antibiotics. This is partly due to the lack of a targeted healthcare regimen that requires pathogen identification. As traditional cultures can take several days, there is a need for a real-time diagnostic tool for identifying uropathogens. In our study, the use of rapid olfactory-based smart sensors for the strain identification was explored by taking measurements of in-vitro assays involving the inoculation of two strains of E. coli (K12 and Uropathogenic E. coli) in urine. Using a Support Vector Machine classifier to train on two previous data collections yielded 95.83 % accuracy when tested on the third subsequent collection. These results demonstrate the promise of smart sensing in differentiating close metabolomic signatures of E. coli and are expected to improve with more extensive data collection, assay development and sensor augmentation.