Application of hyperspectral imaging as a tool to assess microbial responses

Abstract

In the last few years, advances in innovative science have provided new technologies to enhance food safety and quality. Sensitive and rapid microbial detection methods are a key solution for the prevention of foodborne diseases. This has a major ripple effect on health and safety world-wide. Standard microbial detection methods such as direct plate counting, immunological methods and methods based on polymerase chain reactions have the main limitations of being time-consuming and labour-intensive. Research in novel and progressive techniques such as hyperspectral imaging enables the development of new microbial detection methods that can overcome the limitations of conventional methods. This study deals with the simultaneous use of hyperspectral imaging and the conventional plate counting method to assess the growth patterns of Escherichia coli NCTC 12900 (sub-type of E. coli 0157) stored at different environmental temperatures. The growth curves and generation times obtained by analysing microbial data shows that E. coli NCTC 12900 grows better at 400C rather than the characteristic optimal temperature of 370C associated with most E. coli 0157 strains. Hyperspectral images of test and control Agar stimulant systems were analysed using Matlab software to obtain separate normalized histograms. The Bhattacharyya distance analysis was then used to obtain plots which show the pattern of microbial growth with time in relation to light absorbance and reflectance detected by the hyperspectral camera. Due to various limitations and confounding variables in the method used in this study to acquire and assess hyperspectral images, correlation between data obtained in the traditional plate counting method and hyperspectral imaging was not possible. Consequently, further research needs to be carried out to better understand the on-line applicability of hyperspectral imaging in quality control measures related to food quality and safety.