Differentiation of environmental bacteria using hyperspectral imaging technology and multivariate analysis

Abstract

Abstract Efficiency in the identification of bacterial isolates is of paramount importance in clinical microbiology for the correct diagnosis of infections and their subsequent treatment. It is also highly required in environmental microbiology to identify new isolates and the knowledge of the biodiversity of an environment. This raises the need for the development of identification methodologies, which make this practice accessible, safe, and at lower costs, with the emergence, in this sense, of the use of hyperspectral imaging technologies (HSI). In this study, we investigated the potential of HSI in differentiating different bacterial genera. For this, bacterial isolates obtained from water samples were previously sequenced and identified based on the 16S rRNA gene and then grown in Petri dishes containing culture medium. The HSI of each isolate was acquired using an short infrared wave imaging system (SWIR) at the SisuCHEMA workstation. Principal component analyzes showed a clear differentiation between genera and bacterial species. To transform these differentiation assays into an applicable identification methodology, it is necessary to build robust databases capable of covering the wide range of bacterial diversity. The primary results demonstrate that HSI is an objective tool and widely capable of promoting improvements in the identification of bacteria.