Machine learning-enabled non-destructive paper chromogenic array detection of multiplexed viable pathogens on food

Area of Science:

• Food safety
• Microbiology
• Analytical chemistry
• Machine learning applications

Background:

• Rapid and simultaneous identification of multiple viable foodborne pathogens is essential for public health.
• Current methods often require extensive sample preparation, culturing, or incubation, delaying results.

Purpose of the Study:

• To develop and validate a novel pathogen identification system for food safety applications.
• To enable fast, non-destructive, and simultaneous detection of multiple viable pathogens on food matrices.

Main Methods:

• Development of a paper chromogenic array (PCA) using 23 chromogenic dyes and dye combinations.
• Impregnation of the paper substrate with dyes that change color upon exposure to pathogen-emitted volatile organic compounds (VOCs).
• Digitization of color changes and training of a multi-layer neural network (NN) for pathogen identification and quantification.

Main Results:

• The PCA-NN system achieved high-accuracy (91-95%) strain-specific pathogen identification and quantification.
• Successfully distinguished between viable Escherichia coli, E. coli O157:H7, and other viable pathogens.
• Demonstrated simultaneous identification of E. coli O157:H7 and Listeria monocytogenes on complex food matrices like romaine lettuce.

Conclusions:

• The PCA-NN system provides a promising approach for rapid, non-destructive pathogen detection on food.
• This method eliminates the need for enrichment, culturing, or incubation, significantly reducing detection time.
• The technology has the potential to revolutionize food safety monitoring and public health protection.