Enhancing pathogen identification with high background microflora using an artificial neural network-enabled paper chromogenic array sensor approach

Authors: JIA, ZHEN - University Of Florida; LIN, ZHUANGSHENG - University Of Massachusetts; Luo, Yaguang - Sunny; CARDOSO, ZACHARY - University Of Massachusetts; WANG, DAYANG - University Of Massachusetts; FLOCK, GENEVIEVE - Us Army Natick Center; THOMPSON-WITRICK, KATHERINE - University Of Florida; YU, HENGYONG - University Of Massachusetts; ZHANG, BOCE - University Of Florida

Submitted to: Sensors and Actuators B: Chemical
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/19/2024
Publication Date: 3/20/2024
Citation: Jia, Z., Lin, Z., Luo, Y., Cardoso, Z., Wang, D., Flock, G., Thompson-Witrick, K.A., Yu, H., Zhang, B. 2024. Enhancing pathogen identification with high background microflora using an artificial neural network-enabled paper chromogenic array sensor approach. Sensors and Actuators B: Chemical. 410: Article e135675. https://doi.org/10.1016/j.snb.2024.135675.
DOI: https://doi.org/10.1016/j.snb.2024.135675

Interpretive Summary: Technologies for the non-destructive detection of human pathogens are critically needed to ensure food safety and public health. However, the presence of a large population of background microorganisms often interferes with the detection accuracy of targeted pathogens. In this study, we developed and tested a new method using artificial neural network-driven paper chromogenic array sensor technique for the nondestructive, continuous, and simultaneous detection of two major food-borne human pathogens. We demonstrated that the developed method successfully identified pathogens from other indigenous microorganisms with the accuracy reaching 92%. This study benefits food processors with improved technologies for the detection of harmful bacteria on food products.

Technical Abstract: Biohazards, which may occur at all stages of the supply chain, pose significant threats to food safety and public health. Addressing these concerns and enhancing food safety necessitates a nondestructive pathogen surveillance approach capable of continuous and simultaneous detection of multiple pathogens. Detecting and differentiating low concentrations of pathogenic bacteria amid high background microflora levels in foods is challenging, requiring technology with high sensitivity and robust discriminatory capability. This study introduces an artificial neural network -driven paper chromogenic array sensor (ANN-PCA) technique developed for the nondestructive, continuous, and simultaneous detection of Salmonella Enteritidis (SE) and Escherichia coli O157:H7 (Ec) from a high background microflora in shredded cheddar cheese. This method enables accurate detection of SE and Ec in monoculture and cocktail culture, while distinguishing them from a high level of background microflora (~7.5 log CFU/g), with accuracies ranging from 72 ± 11% to 92 ± 3%. In addition, SE and Ec were successfully identified at concentrations as low as 1 log CFU/g within one day, with an accuracy of 72 ± 11%. This approach exhibits promising potential for integration into a digitalized, smart, and resilient nondestructive surveillance system for real-time pathogen detection in foods throughout the supply chain without enrichment, incubation, or other sample preparation steps.