Abstract
Every year, bacterial infections are responsible for over 7 million deaths globally. Timely detection and identification of these pathogens enable timely administration of antimicrobial agents, which can save thousands of lives. Most of the currently known approaches that can address these needs are time- and labor consuming. In this study, we examine the potential of innovative nano-infrared spectroscopy, also known as atomic force microscopy infrared (AFM-IR) spectroscopy, and machine learning in the identification of different bacteria. We demonstrate that a single bacteria cell is sufficient to identify Borreliella burgdorferi, Escherichia coli, Mycobacterium smegmatis, and two strains of Acinetobacter baumannii with 100% accuracy. The identification is based on the vibrational bands that originate from the components of the cell wall as well as the interior biomolecules of the bacterial cell. These results indicate that nano-IR spectroscopy can be used for the nondestructive, confirmatory, and label-free identification of pathogenic microorganisms at the single-cell level.