Campylobacter is an emerging zoonotic bacterial threat in the poultry industry. The current methods for the isolation and detection of Campylobacter are culture-based techniques with several selective agars designed to isolate Campylobacter colonies, which is time-consuming, labour intensive and has low sensitivity. Several immunological and molecular techniques such as enzyme-linked immunosorbent assay (ELISA) and Latex agglutination are commercially available for the detection and identification of Campylobacter. However, these methods demand more advanced instruments as well as specially trained experts. A hyperspectral microscope imaging (HMI) technique with the fluorescence in situ hybridisation (FISH) technique has the potential for multiplex foodborne pathogen detection. Using Alexa488 and Cy3 fluorophores, the HMI (450–800 nm) technique was able to identify Campylobacter jejuni stains with high sensitivity and specificity. In addition, HMI was able to classify six bacteria using scattering intensity from their spectra without a FISH fluorophore. Overall classification accuracy of quadratic discriminant analysis (QDA) method for six bacteria including Bifidobacter longum, Campylobacter jejuni, Clostridium perfringens, Enterobacter cloacae, Lactobacillus salivarius and Shigella flexneri using the HMI technique without fluorescent markers was approximately 88.6 % with pixel-wise classification.

Keywords: acousto-optic tuneable filters, Campylobacter, detection, poultry, food safety, hyperspectral microscope imaging