Nondestructive multiplex detection of foodborne pathogens with background microflora and symbiosis using a paper chromogenic array and advanced neural network

Authors: ZHEN, JIA - University Of Massachusetts; Luo, Yaguang - Sunny; WANG, DAYANG - University Of Massachusetts; DINH, QUYNH - University Of Massachusetts; LIN, SOPHIA - University Of Massachusetts; SHARMA, ARNAV - University Of Massachusetts; BLOCK, ETHAN - University Of Massachusetts; YANG, MANYUN - University Of Massachusetts; GU, TINGTING - University Of Massachusetts; PEARLSTEIN, ARNE - University Of Illinois; YU, HENGYONG - University Of Massachusetts; ZHANG, BOCE - University Of Massachusetts

Submitted to: Biosensors and Bioelectronics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/12/2021
Publication Date: 4/18/2021
Citation: Zhen, J., Luo, Y., Wang, D., Dinh, Q., Lin, S., Sharma, A., Block, E.M., Yang, M., Gu, T., Pearlstein, A.J., Yu, H., Zhang, B. 2021. Nondestructive multiplex detection of foodborne pathogens with background microflora and symbiosis using a paper chromogenic array and advanced neural network. Biosensors and Bioelectronics. 183:113209. https://doi.org/10.1016/j.bios.2021.113209.
DOI: https://doi.org/10.1016/j.bios.2021.113209

Interpretive Summary: Food products contaminated with harmful bacteria such as human pathogens have resulted in many foodborne illness outbreaks. Early detection and subsequent removal of contaminated foods from the supply chain will save lives and support the sustained industry growth. In this current study, we report a novel nondestructive detection of human pathogens in the presence of complex background microflora and symbiosis using an inexpensive paper chromogenic array enabled by an advanced neural network platform. We also validated the detection method using cantaloupe fruits. Research findings benefit pathogen detection method developers, the food industry and consumers.

Technical Abstract: We have developed an inexpensive, standardized paper chromogenic array (PCA) integrated with a machine learning approach to accurately identify single pathogen (Listeria monocytogenes, Salmonella Enteritidis, or Escherichia coli O157:H7) or multiple pathogens (with or without symbiosis) in the presence of background microflora on food. Cantaloupe, a commodity with large and diverse populations of background microflora and volatile organic compounds (VOCs), were used as the model food. The PCA was fabricated from a paper microarray via photolithography and paper microfluidics, and subsequently loaded with 22 chromogenic dye spots and three color standards (Red, Green, and Blue). When exposed to VOCs emitted by pathogens of interest, dye spots exhibited distinguishable color changes and pattern shifts, which were automatically segmented and digitized into a 'R/'G/'B database. We constructed an advanced deep feedforward neural network with a learning rate schedule, L2 regularization, and shortcut connections. After training on the 'R/'G/'B database, the network demonstrated high performance in identifying pathogens in monoculture and in cocktail culture (with or without symbiosis), and in distinguishing them from the background signal on cantaloupe, providing accuracy of monoculture, multiple pathogens, and symbiotic cocktails up to 93% and 91% under ambient and refrigerated conditions, respectively. With its combination of speed, reliability, portability, and low cost, this nondestructive approach holds great potential to significantly advance culture-free pathogen detection and identification on food and is readily extendable to other food commodities with complex microflora.