Hyperspectral imaging for food-related microbiology applications

Abstract Traditional methods for microbial identification are time-consuming procedures that require highly specialized training. These methods are routinely based on phenotypic tests, including Gram staining, culture and growth characteristics, and biochemical analyses. Complete identification can typically take between 5 and 7 days using these tests, and even longer for bacteria that are stressed, difficult to culture, or slow growing. Alternative microbiological techniques, immunoassays, polymerase chain reaction–based methods and biosensors have recently facilitated microbiological characterization, speeding up the analysis up to 1 or 2 days and requiring less specialized expertise. In this context, vibrational spectroscopy, supported by calibrations built with reference data from traditional or alternative microbiological techniques, can provide the means for faster microbiology diagnosis. In addition, this approach allows practical industrial applications: for example, screening of large samples through online implementations. This chapter provides an overview of the state of the art in hyperspectral imaging applications for microbial characterization.